Integrated PET/CT Fusion Imaging and Endoscopic Ultrasound in the Pre-operative Staging and Evaluation of Esophageal Cancer
- 303 Downloads
Accurate staging of esophageal cancer (ECA) is critical in determining appropriate therapy. Endoscopic ultrasound (EUS), computed tomography (CT) and positron emission tomography (PET) scanning can be used, but limited data exists regarding the use of combined PET/CT fusion imaging and EUS in ECA staging. The objective of this study is to evaluate the role of integrated PET/CT imaging and EUS in the staging of ECA.
Identification of patients diagnosed with ECA from 2004 to 2007 that underwent staging PET/CT and EUS. Data regarding tumor detection, lymph node identification, presence of metastatic disease, and affect on patient management were collected and compared between PET/CT and EUS.
Eighty-one patients (65 male, 16 female) were identified with mean age of 63.5 years who underwent EUS and PET/CT to stage known ECA. PET/CT identified the primary tumor in 74/81 (91.4%) of cases, compared to 81/81 (100%) with EUS. Locoregional adenopathy was seen by PET/CT in 29/81 (35.8%) of cases, compared to 49/81 (60.5%) by EUS (p = 0.0001). PET/CT identified celiac axis adenopathy in 8/81 (9.9%) of cases, compared to 11/81 (13.6%) with EUS (p = 0.5050). PET/CT identified 17/81 (21.0%) of patients with distant metastases who subsequently did not undergo attempt at curative surgical resection.
In ECA, EUS is superior to PET/CT for T staging and in identifying locoregional nodes, while PET/CT provides M staging. EUS and integrated PET/CT appear to independently affect treatment decisions, indicating complimentary and necessary roles in the staging of ECA.
Key wordsEndoscopic ultrasound PET/CT Esophageal cancer
Fine needle aspiration
Positron emission tomography
Esophageal cancer (ECA) remains a challenging disease to treat with poor long-term survival and cure rates that have only modestly improved over the last 30 years . Surgery as the sole treatment for advanced locoregional disease has been challenged, and has prompted evaluation of adjuvant and neo-adjuvant therapeutic strategies aimed at improving patient survival . The proper selection of patients for specific treatment regimens may ultimately depend upon the pre-operative assessment of disease stage.
Despite the importance of pre-treatment staging, no one test, or collection of tests, for the staging of esophageal cancer has been accepted as the standard of care. While no one test provides all of the necessary information, current practice varies and may include the use of computed tomography (CT), positron emission tomography (PET), and endoscopic ultrasound (EUS) either alone or in various combinations in the pre-treatment staging process. EUS can provide tumor (T) staging as well as identify local lymphadenopathy. In conjunction with fine needle aspiration (FNA), EUS can also provide a tissue diagnosis of malignant appearing lymphadenopathy. However, because metastatic disease frequently occurs in more distant locations including non-regional lymph nodes, liver, lungs, bone, and adrenal glands, EUS cannot be used for detection of metastatic disease . Metastatic staging has traditionally been performed with CT alone and more recently PET scanning has been studied and employed for the use of esophageal cancer staging. Theoretically integrated PET/CT may be of benefit in staging by combining the ability to identify possible metastases and determine if they are metabolically active in the same test. However, very little data exist evaluating PET/CT in esophageal cancer staging particularly when it is used in conjunction with endoscopic ultrasound.
Thus, the primary aim of this study is to assess the role of PET/CT when used in combination with EUS in the pre-therapeutic staging of esophageal cancer. A secondary aim will be to address the impact that EUS and PET/CT has on management in esophageal cancer.
Materials and Methods
Beginning in May 2004 our institution transitioned from a threefold staging protocol (EUS, CT, and PET) to the present protocol including EUS and integrated PET/CT in all patients with biopsy-proven esophageal cancer being considered for surgical resection. This study was conducted using an electronic patient record system to evaluate selected outcomes of patients seen at the University of Wisconsin Hospital and Clinics, Madison, WI, USA. The study was reviewed and approved by our Institutional Review Board.
EUS and PET/CT studies were performed in all included patients, with the order each study was performed not determined by protocol. EUS staging was completed by the gastroenterologist performing the EUS exam and PET/CT interpretation was performed by a nuclear medicine physician. Both tests were interpreted in a non-blinded fashion as clinical exams. Following pre-operative staging, patient management was recorded, including whether patients underwent primary surgical resection, surgical resection after neo-adjuvant therapy with chemoradiation, surgical therapy followed by adjuvant chemoradiation, chemoradiation alone, or strictly palliative therapy.
EUS and PET/CT Procedure Description
EUS was typically performed as a second endoscopy after the initial endoscopy was performed diagnosing malignancy. The EUS examination was performed with an Olympus GF-UM130 or GF-UM160 radial array echoendoscope (Olympus America, Center Valley, PA, USA) with both 7.5 and 12.0 MHz frequencies. Position (cervical, thoracic, or gastroesophageal junction) and length of tumor in centimeters were recorded. The tumor was evaluated for T stage (T1, invasion into the mucosa or submucosa; T2, invasion into muscularis propria; T3, invasion into adventitia; and T4, invasion into adjacent structures), for the presence of regional lymph node metastases (N0 or N1), and for the presence or absence of celiac axis nodes (CAX). Lymph node EUS features including rounded shape, well-demarcated borders, and a hypoechoic interior were used to determine whether visualized nodes appeared benign or malignant. Larger size (>10 mm) was considered to increase the chance of malignancy, but definitive size criteria of greater than 1 cm was not considered an absolute necessity in the determination of malignancy . Lymph nodes identified on EUS were not reported unless there were features concerning for malignancy. FNA was performed at the time of EUS if the endosonographer felt that (1) performing FNA of a regional or celiac node was technically feasible, (2) if the needle would not traverse the primary tumor, or (3) if the histologic results would alter management. In preparation for PET/CT, patients fasted for a minimum of 4 h prior to the intravenous administration of approximately 10 mCi of 2-deoxy-2-[18F]fluoro-d-glucose (FDG; adjusted for weight). In patients with diabetes, the fasting blood sugar was checked prior to FDG administration to be sure the level was below 200 mg/dl. After the administration of FDG, patients rested quietly for 45–60 min, then were asked to void prior to entering the PET/CT scanner (GE Discovery LS PET/CT scanner). Next, the dose, noncontrast CT scan was acquired with the patient’s arms down at their side (120 mA, 140 kV, 5 mm slices). This CT scan was used for attenuation correction and signal localization of the emission PET scan data. Next, the FDG-PET scan was acquired from the skull base to the thighs for 5 min per bed position. Images were viewed in the transaxial, coronal, and sagittal planes, with images co-registered with the CT scan images when needed.
Determining Change in Management
Prior to staging procedures with EUS and PET/CT, all included patients were initially considered surgical candidates. Surgery was thus considered the baseline treatment and any variation in this decision as a result of a staging test was considered a change in management. At our institution, patients with locally advanced disease (T3 N0 or T1–3 N1) are recommended to undergo neo-adjuvant therapy with chemoradiation followed by surgery, which would signify a change in management based on pre-treatment staging. Similarly, the presence of distant metastases or invasion into local structures thereby prohibiting a patient from potential curative surgical resection was also considered a change in management. Patients who ultimately did not undergo surgery because of reasons other than pre-operative imaging, particularly patients with co-morbid illness that eventually precluded surgery, were not included in the data analysis of impact on management (N = 12).
Determining Tumor and Nodal Staging Accuracy
Pathology staging data was obtained for all patients (N = 24) who underwent primary surgical resection without antecedent chemoradiation therapy that could affect the stage of the cancer. The pathologist’s staging of the tumor and nodes in these patients represented the gold standard for which to compare staging with EUS and PET/CT.
The two imaging modalities were compared for their classification of the locoregional adenopathy and celiac axis nodes into normal and abnormal categories. The difference in the percent success in the modalities was assessed using McNemar’s test.
Patient and Tumor Characteristics
The study population consisted of 81 patients [65 (80.2%) men; 16 (19.8%) women] with biopsy-proven adenocarcinoma (N = 64; 79.0%) or squamous cell carcinoma (N = 17; 21.0%) of the esophagus diagnosed from May 2004 to December 2007; a 41-month period. The mean (±SD) age was 63.5 (±11.6) years. The mean number of follow-up months was 11 months. The tumor was located in the cervical esophagus (N = 3; 3.7%), the thoracic esophagus (N = 19; 23.5%), and at the gastroesophageal junction (N = 59; 72.8%). Mean tumor length was 5.8 cm.
Celiac Axis Nodes
Change in Management
Based on the abovementioned criterion for determining change in management, and after excluding the 12 patients that did not undergo surgery for other reasons, PET/CT detected metastatic disease and directed the care to either chemoradiation therapy or palliative care in 17 of 69 cases (24.6%). EUS re-directed patient care to neo-adjuvant therapy prior to surgical resection in 26 of 69 cases (37.7%). Among these 26 patients who underwent neo-adjuvant therapy due to EUS staging, six had nodal involvement or locoregional disease on PET/CT. Thus, EUS improved the ability to provide locoregional staging in an additional 20 patients as compared to PET/CT.
Tumor and Node Staging Accuracy
Among the subset of patients that underwent surgery without antecedent chemotherapy (N = 24), the surgical pathology staged seven T1 (29.2%), four T2 (16.7%), and 13 T3 (54.2%) and EUS staged six T1 (25.0%), seven T2 (29.2%), ten T3 (41.7%), and one Tx (4.2%). A comparison within this patient subset showed that EUS was accurate in T staging in 18 of 24 (75.0%) of cases.
With regards to nodal staging accuracy of EUS and PET/CT compared to the surgical pathology in patients who underwent primary resection (N = 24), there were 12 cases without regional lymph node metastasis and 12 cases with regional lymph node metastasis. Nodal staging by EUS agreed with the surgical pathology data in 15 of 24 cases (62.5%). Nodal staging by PET/CT agreed with the surgical pathology data in 13 of 24 cases (54.2%).
Studies of esophageal cancer staging with EUS have demonstrated strength in T and N staging accuracy , prediction of patient survival , and cost-effectiveness . However, a significant weakness of EUS staging alone is that it cannot provide the most important factor in staging—the presence or absence of metastatic disease. There is limited data available regarding the use of separate PET and CT for esophageal cancer staging and even less data on combined PET/CT, especially concerning pre-treatment staging and its effect on patient management [8, 9, 10]. To our knowledge, no study has examined a staging protocol combining integrated PET/CT and EUS in pre-operative and pre-treatment staging esophageal cancer. Therefore, it was the primary aim of our study to assess the complimentary roles of EUS and PET/CT in the staging of esophageal cancer. In addition, we aimed to assess how the staging results from EUS and PET/CT were being used clinically to affect management.
While surgery has traditionally been the treatment of choice for esophageal cancer , the presence of lymph node involvement and metastatic disease at the time of diagnosis makes resection alone inadequate for all patients and may portend a poor long-term prognosis . Therefore, pre-treatment TNM staging is an important determinant of deciding who would benefit from surgical resection alone and who may benefit from possible neo-adjuvant therapy. Our results further validated previous studies suggesting that EUS is the superior locoregional staging modality . In our study, EUS was able to identify the primary tumor in every instance and was able to provide T staging information in all but one instance. Additionally, EUS identified more patients with regional lymphadenopathy (N staging) when compared to PET/CT (60.5% vs. 35.8%). Therefore, it appears even with integrated PET/CT, EUS is still the superior imaging modality for locoregional lymphadenopathy.
However, as EUS cannot reliably detect distant metastases, PET/CT proved invaluable in this respect identifying 17 patients with metastatic disease. In addition, PET/CT allowed visualization of the celiac axis and investigation for celiac axis nodes in the 15 patients where the celiac axis could not be assessed by EUS because of inability of the EUS scope to traverse the stenotic esophageal cancer. Thus, the use of EUS and PET/CT provided complimentary information improving pre-treatment staging beyond their use alone.
The importance of pre-treatment staging may be further strengthened if those who would receive benefit from neo-adjuvant chemotherapy can be identified. To date, there have been both positive [14, 15, 16, 17, 18] and negative [19, 20] studies of the utilization of neo-adjuvant chemoradiation therapy. A Cochrane review  of 11 randomized trials demonstrated a 12% survival advantage for pre-operative chemotherapy compared with surgery alone. This is an important point suggesting pre-treatment staging may change or alter management. At our institution, patients identified as having a T3 or N1 lesion are recommended to undergo neo-adjuvant therapy with chemoradiation prior to surgery, while the presence of distant metastases or invasion into local structures would also constitute a necessary change in management from attempting curative surgical resection. In our study, EUS identified the need for neo-adjuvant therapy, or change in management, in 26 of 69 cases (37.7%).
Combined PET/CT imaging identified presence of distant metastatic disease including most commonly distant lymphadenopathy as well as liver and lung lesions and even a metastasis to muscle. These findings led to a change in management by precluding esophagectomy in 17 of 69 cases (24.6%). Yet overall, the percentage of cases that PET/CT changed management is similar to our previously published data showing that separate diagnostic CT and PET scans changed management in 25% of patients. This suggests that the performance of PET and CT, either as a separate test or in a single fused test, increases the number of patients identified with metastatic disease. PET/CT in our study did show a greater clinical impact than in Bar-Shalom et al. who previously demonstrated combined PET/CT had changed patient management in only 10% of patients by detecting nodal metastases that warranted upstaging and by excluding disease in sites of benign uptake after surgery as compared with PET alone .
In our previous study of separate staging with CT and PET and EUS , 34.8% of patients had their management changed from direct surgical treatment to neo-adjuvant therapy based on EUS staging and 72.3% of patients had management changed because of combined pre-operative staging using the separate individual tests of EUS, CT, and PET. Vazquez-Sequerios et al.  examined the value of EUS staging and EUS-FNA in affecting management and found that the combination of EUS and CT changed management from direct surgical therapy in 77% of patients by resulting in pre-operative chemoradiation or non-surgical treatment therapies. Similarly, we found in our current study that when using PET/CT and EUS, 76.6% of the total patient population had their management changed by combined staging with EUS and PET/CT.
It should be noted that neither EUS nor PET/CT are perfect tests with complete accuracy and even when the PET and CT images are fused staging of locoregional adenopathy may be challenging. The staging accuracy of EUS has been well-documented by multiple previous studies to be approximately 61–76% for T staging and 55–90%  for N staging depending on whether the lesion is traversable [25, 26, 27], which is comparable to the accuracy we found in our study in patients who underwent surgery without neo-adjuvant therapy. Currently, there is a paucity of PET/CT staging accuracy data; however, our study determined nodal staging accuracy of PET/CT to be 52.0% in our patient subset (N = 24) which again likely reflects the difficulty in identifying locoregional adenopathy even with fused PET/CT imaging as it is difficult to separate a locoregional node from a primary tumor. However, the primary purpose of our study was not to directly compare the sensitivity and accuracy of EUS and PET/CT in staging esophageal cancer but rather to examine the relative role of EUS and PET/CT when used together in conjunction for staging esophageal cancer.
To our knowledge, this is the first study to evaluate the complimentary roles of EUS and combined PET/CT in the pre-treatment staging of esophageal cancer. Our results have further validated that multimodal screening with locoregional staging provided by EUS and metastatic staging provided by PET/CT is mandatory in esophageal cancer as both modalities heavily influence treatment decisions. We recommend continued application of EUS; however, with PET/CT or separate PET and CT as the primary test performed to rule out metastatic disease. It is unclear from our study if combined PET/CT offers significant advantages over separate PET and CT imaging as our ability to detect adenopathy and metastases and change management was not significantly better with PET/CT than previous reports when CT and PET were used as individual imaging studies for staging esophageal cancer. Further studies are needed to compare accuracy of PET/CT versus separate PET and CT in staging esophageal cancer, the cost-effectiveness of multimodality pre-operative staging for esophageal cancer, and how multimodality staging affects patient outcome.
- 5.Rosch T (1995) Endosonographic staging of esophageal cancer: a review of literature results. Gastrointest Endosc Clin North Am 5:537–547Google Scholar
- 8.Cerfolio RJ, Bryant AS, Ohja B, Bartolucci AA, Eloubeidi MA (2005) 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 129:1232–1241CrossRefPubMedGoogle Scholar
- 16.Bonnetain F, Bouche O, Michel P et al (2006) A comparative longitudinal quality of life study using the Spitzer quality of life index in a randomized multicenter phase III trial (FFCD 9102): chemoradiation followed by surgery compared with chemoradiation alone in locally advanced squamous resectable thoracic esophageal cancer. Ann Oncol 17:827–834CrossRefPubMedGoogle Scholar
- 21.Malthaner RA, Collin S, Fenlon D (2006) Preoperative chemotherapy for resectable thoracic esophageal cancer. Cochrane Database Syst Rev 3:CD001556Google Scholar
- 24.Rosch T, Classen M (1999) Staging esophageal cancer: the Munich experience. In: Van Dam J, Sivak MV (eds) Gastrointestinal endosonography. WB Saunders, LondonGoogle Scholar