The effect of anastomotic leakage on the incidence of recurrence after tri‐modality therapy for esophageal adenocarcinomas

Neoadjuvant chemoradiotherapy (nCRTx) reduces the incidence of recurrence, while anastomotic leakage has shown increase the risk of recurrence. The primary objective of this retrospective study was to investigate the incidence and pattern of recurrence and secondary median recurrence‐free interval and post‐recurrence survival in patients with and without anastomotic leakage after multimodal therapy for esophageal adenocarcinoma.


| INTRODUCTION
Esophageal cancer is currently the eight most common cancer, and is the sixth most common cause of cancer related deaths, worldwide. 1 For patients with locally advanced esophageal cancer curative treatment consists of neoadjuvant chemo(radio) therapy followed by esophagectomy with lymphadenectomy and immunotherapy in R0 ypT+ patients. 2 Patients can be operated by a transhiatal or transthoracic approach, and the resection can be performed either open, minimally invasive or hybrid. 1 After resection of the esophagus, gastrointestinal continuity is usually restored with an esophagogastric anastomosis. Anastomotic leakage is a severe complications and is associated with increased postoperative morbidity and mortality. 3 A number of studies report that anastomotic leakage increases the risk of recurrences. 4,5 This increased risk may be attributed to cell-mediated immunity, cytokines expression, and inflammatory responses, caused by anastomotic leakage. 4,5 Theories suggest that surgery on itself causes a suppression of cell-mediated immunity, and the exfoliation of cancerous cells into the bloodstream, increasing the metastatic potential. For esophageal cancer surgery, it has been shown that the severity of depressed monocyte and T-lymphocyte cytokine immune responses is related to the extent of the surgical trauma. 6,7 This illustrates the multifactorial nature of the development of recurrence after curative cancer treatment.
In other studies, the promoting effect of inflammation on recurrence is supported, for example in breast cancer surgery it is found that patients experiencing postoperative wound complications have a greater risk of developing systemic recurrences. 8 In esophageal cancer patients, anastomotic leakage was found to be an independent risk factor of cancer recurrence. 9,10 Retrospective research on the impact of severe esophageal anastomotic leakage, defined as a symptomatic disruption classified as grade III or IV according to the Clavien−Dindo classification, on long-term survival and recurrence reported a significant association between severe esophageal anastomotic leakage and loco-regional and combined (coexisting loco-regional and distant) recurrences, without an association with distant recurrences. 10,11 In this study only 28.6% of patients received neoadjuvant chemoradiotherapy (nCRTx). The use of nCRTx did not affect the incidence of anastomotic leakage, as described by several authors, however, it did decrease the incidence of locoregional recurrence. 12,13 The impact of experiencing anastomotic leakage in the postoperative course on cancer recurrence after esophageal cancer surgery in the current era of nCRTx is unknown.
Therefore, the aim of this study was to investigate what the effect of anastomotic leakage was on the incidence of recurrent disease and recurrence patterns in patients after treatment with trimodality therapy with curative intent (nCRTx followed by an esophagectomy) for esophageal cancer (Central Illustration 1).  14 Data was extracted from the institutional prospectively maintained databases containing data on esophagectomy patients for esophageal and gastroesophageal cancer. Patients were provided the opportunity to opt-out if they did not agree with anonymous use of their medical data for research purposes. Ethical approval for the data used for this study was waived by the local institutional review board. The STROBE guidelines were followed to ensure correct reporting of the study methods and results. 15

| Patient population
Consecutive patients who underwent an esophagectomy, for a histologically proven, resectable (cT1-4a, N0-3, M0) primary adenocarcinoma of the esophagus or gastroesophageal junction, between January 2010 and December 2018, were eligible for inclusion. Inclusion

| Outcome measures
The primary outcome was the incidence of recurrent disease in patients who experienced anastomotic leakage compared to patients without anastomotic leakage in the postoperative course, specified for different sites (loco-regional, distant, and combined loco-regional and distant) after curative multimodal treatment. The recurrence free interval, defined as the median time in weeks between surgery and confirmation of recurrence, was calculated for the different recurrence sites. Additionally, the post-recurrence survival, defined as interval in weeks from date of confirmation of the recurrence to the date of death, was calculated, for the total study population and for patients specified to recurrence site groups. Follow-up on survival was collected until August 2021. For post-recurrence survival at least 12 months follow-up after recurrence was chosen for analysis, since literature describes a median survival after recurrence of 3−10 months. [16][17][18] The secondary outcomes were: (1) the specific location of the recurrent disease specified for patients with and without anastomotic leakage, and (2) location of recurrence with regard to the radiation fields during the neoadjuvant radiotherapy.

| Treatment and follow-up
Patients included in the current study successfully completed at least 80% of the intended nCRTx, followed by an esophagectomy. The extent of the lymphadenectomy consisted of a total mediastinal two-field lymph node dissection in one center (AMC), and a standard two-field

| Definitions of anastomotic leakage confirmation and severity
Anastomotic leakage confirmation was defined by the following methods or when the following signs and symptoms were present: (1) radiologically: a contrast swallow showing contrast medium extravasating from the anastomosis or intrathoracic stomach into the mediastinum or a CT-scan showing mediastinal fluid and air, anastomotic wall discontinuity, or a fistula, (2) endoscopically: signs of conduit necrosis or full thickness defect in de esophagogastric wall, (3) physical examination: signs of an esophago-cutaneous fistula, with saliva draining from the cervical wound in case of a cervical anastomosis. 19 The severity of the anastomotic leakages was scored according to the Esophagectomy Complications Consensus Group (ECCG) classification (type I: local defect requiring no change in therapy or treated medically or with dietary modifications, type II: localized defect requiring interventional but not surgical therapy, for example, interventional radiology drain, stent, or bedside opening, and packing of incision, and type III: localized defect requiring surgical therapy). 20

| Definitions and locations of recurrence
The recurrence date was the date of pathological confirmation of the recurrence or the date of a strong suspicion of recurrence if pathological confirmation was not possible or not indicated (poor clinical condition that would make treatment impossible or patient preference). Disease recurrence sites were classified as "loco-regional," "distant," or a combination of loco-regional and distant ("combined loco-regional and distant"). Loco-regional recurrences were defined as located at the site of the primary tumor location and/or in loco-regional lymph nodes, at the anastomotic site or in the gastric conduit/esophagus. Distant recurrences were defined as systemically located, or located in distant lymph nodes outside the surgical or radiation field.
Combined recurrences were defined as a combination of locoregional and distant recurrence. The radiation oncologist recorded whether the recurrence was in-or outside of the targeted area for radiotherapy. Frequencies and percentages were given for recurrent disease specified for each recurrence site in patients who experienced anastomotic leakage. The median in weeks for the recurrence free interval was analyzed for each recurrence site (loco-regional, distant, and combined loco-regional and distant) and specified for the occurrence of anastomotic leakage, using Kaplan−Meier curves and log-rank tests.

| Patients with anastomotic leakage in the postoperative course
Most leakages were confirmed radiologically (n = 77; 84.6%), followed by endoscopically (n = 7; 7.7%), or by physical examination (n = 7; 7.7%). Details on the severity of the anastomotic leakage according to the ECCG classification were as follow, 8 patients (8.8%) had a type I, 51 (56%) had a type II, and 32 (35.2%) had a type III. Information on treatment of the leakage is summarized in Table 2.
Distant recurrences were most frequently found in the liver (35.1%) and bone (32.4%). Loco-regional recurrence were most often found in loco-regional thoracic lymph nodes (42.9%) or located at the anastomosis and gastric conduit (23.8%) (Supporting Information: Table 1). The location of the recurrent disease with regard to the radiation field is displayed in Table 4.

| Recurrence free interval
Overall, the median recurrence free interval was 39 weeks (95% CI:

| Post-recurrence survival
Overall, the median post-recurrence survival for patients with anastomotic leakage was 11 weeks (95% CI: 6   Reconstruction with pedicle flap 4 3 7 Resection and temporary esophagostomy 3 2 5 Abbreviation: ECCG, Esophagectomy Complications Consensus Group. a The total of patients for subgroups, will not add up to the total number of patients in the study, as multiple treatments may apply for 1 patient. leakage in the postoperative course (p = 0.702; Figure 6). The majority of the patients received either palliative intent therapy (n = 20; 45.5%) or best supportive care (n = 22; 50%), only 2 patients (4.5%) received tumor directed therapy with curative intent ( Table 5).

| Discussion
This retrospective cohort study investigated the influence of anastomotic leakage on the incidence and pattern of recurrent disease in patients after curative multimodal therapy for esophageal cancer. The results show, that patients with anastomotic leakage did not experience recurrence more frequently than those without this complication in the postoperative course. No statistical differences were found in the incidence of the occurrence of loco-regional, distant or combined loco-regional, and distant recurrence between patients with and without anastomotic leakage. Patients with anastomotic leakage that developed recurrent disease did have a significantly shorter recurrence free interval compared to patients without this postoperative complication.
In colorectal surgery, it is assumed that the oncological effect of anastomotic leakage involves the release of cancer cells that stayed behind in the bowel lumen. [21][22][23][24] Some studies found exfoliated tumor cells in irrigation fluid from the rectal stump or stapling devices. 25,26 In esophageal cancer, it is also suggested that the spillage of tumor cells plays a role in the development of recurrence after anastomotic  27,28 It is assumed that this inflammatory environment enhances mostly loco-regional recurrences. The findings in the current study did not mirror the findings of previous studies. This contrasting result compared to previous studies might be explained by the fact that this study is the first to include only patients that received nCRTx, which is currently part of the gold standard for the treatment of patients with resectable esophageal cancer in the Netherlands. It has been proven that this neoadjuvant regimen alters the recurrence pattern and results in less loco-regional recurrences. 29-31 More research is needed, but this might suggest that F I G U R E 3 Recurrence free interval in patients with a loco-regional recurrence specified for patients with and without experiencing anastomotic leakage postoperatively.
F I G U R E 4 Recurrence free interval in patients with a distant recurrence specified for patients with and without experiencing anastomotic leakage postoperatively.
the neoadjuvant radiotherapy affects the tissue to such an extent that fewer in-field recurrences occur. This could explain why anastomotic leakage in this study did not influence the incidence of loco-regional recurrences significantly. This idea is substantiated by the fact that in this study also the incidence of combined recurrences and the post-recurrence survival did not differ significantly between patients with or without anastomotic leakage. It is possible that after radiotherapy the tissue surrounding the tumor and future F I G U R E 5 Recurrence free interval in patients with a combined loco-regional and distant recurrence specified for patients with and without experiencing anastomotic leakage postoperatively.  This may result in a larger difference in the occurrence of recurrence and recurrence free interval.
The overall goal should however be to strive for less postoperative anastomotic leakage, new methods for prevention of anastomotic leakage such as ICG-fluorescence angiography, may provide relief and help reducing the incidence of leakage. 33,34 Several strengths and limitations of this study should be acknowledged. A major strength of this study is that all included patients were treated with the same nCRTx regimen. Therefore, these outcomes are generalizable to patients treated with curative intent in the current era of nCRTx. Another strength is that the details of the recurrence sites were available for all patients, providing knowledge of the incidence and specific location of recurrent disease. A limitation is the retrospective nature of this study, therefore only a correlation and no causality could be proven.
Furthermore, due to a small sample size our outcomes could be biased and differences could potentially be explained by possible confounders within this population that were not studied.

| Conclusion
This retrospective cohort study shows that experiencing anastomotic leakage after curative intent treatment with nCRTx and esophagectomy does not result in a higher incidence of loco-regional, distant or combined loco-regional, and distant recurrences. Patients after anastomotic leakage do have a significantly shorter recurrence free interval. This may have consequences for the follow-up. These data support the evidence that neoadjuvant radiotherapy can be protective for the development of loco-regional recurrences and suggest the same protective role in the subgroup with anastomotic leakage.

DATA AVAILABILITY STATEMENT
Data sharing is not applicable to this article as no new data were created or analyzed in this study.