Staging Techniques for Carcinoma of the Esophagus

Esophageal Cancer Tnm Staging

The current staging system for esophageal cancer follows the American Joint Committee on Cancer (AJCC) Cancer Staging Manual, 7th edition, which uses the TNM system of T for tumor, N for regional nodal status, and M for presence or absence of distant metastases ( Table 37-1 ). The TNM staging system was recently revised, and significant changes included modifying the T stage, incorporating the grade of the tumor into the stage, subdividing the N stage based on the number of involved lymph nodes, and separating adenocarcinoma from squamous cell carcinoma (SCC). There is also only one class for metastatic lesions. M1a and M1b, separating distant nodal and distant organ metastases, have been removed. High-grade dysplasia is now T in situ (Tis), and T4 has been divided into T4a and T4b. T4a cancers are locally invasive but are still deemed potentially resectable albeit with a portion of pericardium, pleura, or diaphragm. T4b tumors invade the trachea, aorta, or vertebral bodies and deem the tumors unresectable.

Another major modification in the staging system is the subset of nodal stage. As has been modified for gastric, colon, melanoma, and other malignancies, the number of lymph nodes involved has become a prognostic marker and thus incorporated into the formal staging system. Accurate clinical nodal staging of esophageal cancer is challenging, and typically the patient is clinically staged as either node-positive or node-negative; however, this change in nodal staging is an important prognostic adaptation. Multiple studies from different esophageal groups have found significant survival differences based on absolute number of positive lymph nodes, absolute number of negative lymph nodes, and ratio of positive to negative lymph nodes.

The grade of the tumor is also now a part of the staging classification as poorer differentiation corresponds with worse prognosis. T1 is divided into T1a for intramucosal cancers and T1b for submucosal lesions. This T stage modification is based on the incidence of lymph node involvement and subsequent survival differences for tumors of increasing depth. Other changes include the addition of tumor location to the SCC staging and the staging of tumors of mixed histology as per the SCC group.

Endoscopic Ultrasonography

EUS staging of esophageal cancers was first advanced in the literature by Lightdale in 1992, and in the next decade the modality became a standard part of staging newly diagnosed esophageal cancer patients. Before EUS, CT scan was the primary modality available to decide whether to proceed with an esophagectomy on an otherwise medically fit patient presenting with esophagus cancer, although MRI and bone scans were sometimes indicated by clinical presentation. Now EUS is available to help determine locoregional stage of esophagus cancer as well as distant disease by providing cytology of liver and adrenal or celiac lymph node metastases.

For patients without evidence of distant disease on PET-CT, EUS is used for T and N stages to allow referral for neoadjuvant therapy. It is most often performed by the gastroenterologists; however, some surgical groups are experienced and it would behoove thoracic surgeons to be proficient so that they can offer one-stop shopping for staging esophageal and lung cancer patients.

Conventional EUS is carried out with conscious sedation using the 360-degree radial mechanical echoendoscope with a 7.5-MHz ultrasound probe. Conventional radial EUS is the best approach for assessing depth of primary tumor into the esophageal wall. For EUS-guided fine-needle aspiration (FNA) of suspicious lymph nodes, a curved linear array echoendoscope is used with a 22- or 25-gauge needle. A 7.5-MHz or lower frequency endoscope is better than the high frequency endoscope for assessing regional lymph nodes because of the greater depth of visualization. Endoscopic criteria for potentially malignant lymph nodes include at least two of the following characteristics: round, discrete, hypoechoic, or a dimension greater than 1 cm. EUS imaging of nodes alone is not specific enough, but FNA of the suspicious nodes can result in a specificity rate of up to 95%.

EUS remains the most accurate modality for determining T stage with accuracy rates ranging from 64% to 80% for the low-frequency probe and up to 85% to 92% for the high-frequency probe. Accuracy can only be calculated when patients have undergone histologic confirmation of their tumor and nodal status. The accuracy of EUS for staging of T and N in the most recent series of patients in the past 15 years is summarized in Table 37-2 . As summarized, EUS staging is indicated before initiation of neoadjuvant therapy.

EUS has an accuracy range of 80% to 90% for T3 and T4 stages but for T1 and T2 the rate drops to 74%. In a large series of more than 200 patients with pathologic confirmation of their tumor and nodal stages, most EUS staging errors were understaging of T0 and T1 and overstaging of T2. The reliability of EUS for accurately staging T2N0 tumors was examined by Crabtree and colleagues, who retrospectively studied the Society of Thoracic Surgeons (STS) General Thoracic Surgery Database (GTDB) and concluded that clinical staging of T2N0 tumors is unreliable. They reported a 26% downstaging and 47% upstaging of clinical T2N0 (cT2N0) patients treated with primary esophagectomy. EUS accurately staged T3 and T4 lesions in 85% of cases. In another series, of 47 patients undergoing surgical resection for adenocarcinoma or SCC, EUS overstaged T stage in 19% of patients and understaged it in 17%.

In studies with pathologic confirmation, the accuracy rates for N stage range from 70% to 86%. Before FNA-guided biopsies of lymph nodes by EUS was applied, the false-negative rate of EUS for nodal stage exceeded 33%. Endobronchial ultrasound (EBUS) has also been combined with EUS to improve the staging precision of ultrasound modalities. As demonstrated in a prospective clinical trial (NCT01038544), patients with newly diagnosed esophageal cancer underwent radial and convex EUS followed by convex EBUS to evaluate the standard nodal stations. EBUS complemented the examination of stations 2, 4, and 7 by EUS but also provided access to lymph node level 10R. EBUS added to the sensitivity of the EUS by upstaging 12% of the patients and contributing to the EUS stage in 69% of the patients. The EBUS scope also could be passed across tight malignant strictures, which precluded staging with the EUS but allowed for a clinical T and N stage of tumors that could not otherwise have been staged.

Although EUS has added much to the clinical staging and management of esophageal cancer patients, it still is limited by malignant stenoses preventing staging of the primary tumor. In these cases, however, tumor obstruction almost always represents T3 or T4Nx. In addition, the probability of a T3 or T4 tumor being N1 exceeds 80%.

With the continued expansion of endoscopic therapy for patients with early esophageal cancers, accurate staging of T1 adenocarcinoma with EUS will be important to determine which patients may be offered endoscopic therapy with potential cure and which patients should undergo esophagectomy. From experienced EUS groups, the accuracy of staging an intramucosal (T1a) cancer was 82% to 94%. A T1b cancer has about a 20% likelihood of lymph node metastases versus the intramucosal lesion, which is less than 5%; thus, the EUS T stage may help in choosing between endoscopic and surgical resection. EUS before endoscopic resection of T1 tumors remains controversial, however, because it is not highly accurate and the patient should probably just have the EMR for staging the superficial lesion.

For the patient initially presenting with a localized cancer, conventional EUS with a 7.5- to 12-MHz radial ultrasound probe is good for initial assessment of tumor infiltration and looking for suspicious lymph nodes (hypoechoic, >1 cm). The higher frequency 12- to 20-MHz probe may provide the most accurate staging for the T1 or T2 suspicious tumors, but generally with EUS, pT1 tumors are understaged by 18% and pT2 tumors are overstaged by 47%. The exam should include a full evaluation of the liver and lymph nodes in the paragastric, subhepatic, and celiac stations. If a suspicious node is identified, a curvilinear echoendoscope allows FNA of the suspicious node.