Distribution of lymph node metastases in esophageal carcinoma [TIGER study]: study protocol of a multinational observational study

In order to be eligible to participate in this study, a subject must meet all of the following criteria:

A potential subject who meets any of the following criteria will be excluded from participation in this study:

Preoperative staging will be performed according to national guidelines. This therefore, may differ per country. Usually, patients are staged with an endoscopy with biopsies, an endoscopic ultrasound, a PET-CT-scan of neck, thorax and abdomen and an ultrasound of the neck. An EBUS-TBNA will be performed if indicated. All patients are discussed in a multidisciplinary team (MDT).

A suspected lymph node is defined as a node larger than > 9 mm short axis or a node 5–9 mm short axis that is round, inhomogeneous and has an irregular border (2 out of 3). In case of PET-CT-scan, pathological nodes are those with FDG uptake, a short axis > 10 mm or those with a short axis between 5 and 9 mm that present with sharp borders.

In case of potentially curative disease (cT1-4a, N0–3, M0) patients may be treated with neoadjuvant therapy, this however, may differ per country and usually consists of chemotherapy or chemoradiotherapy.

One coronal, 1 sagittal and 3 axial CT-images of the radiation field is documented in the medical file, to assess lymph nodes stations incorporated in the radiation field. An anatomical description of the radiation field will be given according to highest and lowest radiated lymph node station. An in-field lymph node is defined as a lymph node within the clinical target volume. Clinical target volume fields as defined gross tumor volume plus a margin for sub-clinical disease spread.

After completion of neoadjuvant therapy patients will generally be restaged with a (PET-)CT scan to exclude distant metastases before patients are scheduled for surgery. In addition, the locoregional lymph nodes will be evaluated. The (PET-)CT will usually be performed 2–4 weeks after completion of neoadjuvant therapy.

If no metastases are detected, patients will be operated 5–12 weeks after completion of neoadjuvant therapy. This may differ per country and may depend on the type of neoadjuvant therapy. If no neoadjuvant therapy is administered patients are directly scheduled for surgery.

An esophageal resection with a 2- or 3-field lymphadenectomy can be performed via a thoracolaparoscopy, a thoracolaparotomy or a hybrid procedure and a cervical incision as indicated. A gastric tube or colonic interposition can be used for reconstruction. Lymph node stations may be dissected ex vivo (after enbloc resection with the specimen), in vivo, or partially ex and in vivo, and will be separately sent for pathological examination. Lymph node stations in close proximity to the tumor are preferably marked with sutures or beads to prevent damage to the specimen and circumferential resection margin.

Different classification systems are used in different countries to classify lymph node stations around the esophagus: the AJCC 8th edition esophageal cancer staging and the JES 11th edition esophageal cancer staging. In Fig. 1 and Table 1 these classifications have been combined for the purpose of this study.

In case of a 3-field lymphadenectomy stations 1–19 will be resected and in case of a 2-field lymphadenectomy stations 6–19 will be resected as usual. If not all lymph node stations are resected, patients can still be included in the study, however this has to be documented accurately. A video appendix and definitions of the individual lymph node stations are displayed and available for review on the TIGER study website.

The esophageal resection and lymphadenectomy specimens will be processed and analyzed according to national and international guidelines by the department of pathology. In addition to the separately sent lymph node stations, the esophageal surgical resection specimen will be carefully analyzed for retained lymph nodes and structures macroscopically suspicious for lymph nodes will be embedded. All lymph nodes under 5 mm will be totally embedded for microscopic evaluation, larger lymph nodes will be totally embedded in slices of 3–4 mm thick. Microscopically, a circumscript area of lymphoid cells containing a follicular architecture and/or a subcapsular sinus is identified as a lymph node. In the final report the exact localization and number of lymph nodes found will be reported.

Initial microscopic evaluation will be performed by standard H&E staining. In case of suspicion of micro-metastasis (0.2–2.0 mm) or isolated tumor cells in the lymph node, or in case of suspicion of residual tumor cells in patients with extensive response to neoadjuvant therapy, additional keratin stains will be performed. A metastatic lymph node is defined as a lymph node with tumor cells. Lymph nodes containing micro-metastasis or isolated tumor cells are also considered as metastatic lymph nodes but will be also recorded separately. Lymph nodes with regression after chemoradiation with isolated tumor cells will be considered as metastatic lymph nodes, however, these also will be recorded separately, so that the prognostic value of these isolated tumor cells can be determined. The same applies to fibrotic lymph nodes without vital tumor cells. These are so called preneoadjuvant therapy N+ nodes. In case of an adenocarcinoma of the esophagus, Her2-status will be analyzed (immunohistochemistry and SISH).

Patients will be followed up until 5 years after the operation. Follow-up will usually be scheduled (but differences between countries exist) every 3 months the first year, every 6 months the second until the fourth year and once yearly until the fifth year. Investigations are performed according to national guidelines.

Numbers and percentages of resected lymph nodes and lymph node metastases will be given per lymph node station (Fig. 1). Tumor location and invasion depth will be categorized. Patients with adenocarcinoma and squamous cell carcinoma and patients with and without neoadjuvant therapy will be analyzed separately.

The sensitivity, specificity, and positive and negative predictive values of EUS and PET-CT will be reported. Perioperative morbidity and mortality will be summarized descriptively. For each patient group (squamous cell versus adenocarcinoma, with and without neoadjuvant therapy), explorative cluster analyses will be performed to identify subgroups of patients with different patterns of lymph nodes metastases. Potentially relevant characteristics at the time of surgery like age, gender, tumor location, tumor invasion depth, tumor differentiation, vaso-invasive growth will be included in the analysis. No restrictions will be applied to the number of clusters in each analysis, but the ratio of the largest cluster size to the smallest cluster size should preferably not exceed the value of 3 and/or the smallest cluster size should be minimally 30 patients. Characteristics introducing patient outliers will be excluded and one should further be able to attribute meaning to the resulting cluster profiles. Clusters that show the phenomenon of skip metastases will be noted. The resulting clusters will be evaluated for the diffusion pattern of future metastases during follow-up (descriptive analysis), the number of future metastases during follow-up (Poisson regression or generalized estimation equation, whichever appropriate), for 3- and 5-year overall and disease-free survival (Kaplan-Meier survival analysis). Multivariate analysis will be performed using the Cox hazard regression method. The univariate analysis, including all baseline parameters, will serve as the basis for the multivariate Cox hazard regression model. Variables showing association (p < 0.10) with survival in univariate analysis will be included in the multivariate analysis. Age and sex will be included in all multivariate analyses. Results are presented as hazard ratio with exact 95% confidence interval (95% CI). After 5-years of follow-up the efficacy index will be determined (incidence of metastases to an area (%) × 5-year overall survival rate (%)). A log-rank test, Mann-Whitney U test, or χ²-test will be used as indicated to compare groups. A value of p < 0.05 will be considered statistically significant.