==== Front World J Surg World J Surg World Journal of Surgery 0364-2313 1432-2323 Springer International Publishing Cham 37103558 7013 10.1007/s00268-023-07013-5 Original Scientific Report C-Reactive Protein as a Negative Predictive Marker for Anastomotic Leakage After Minimally Invasive Esophageal Surgery http://orcid.org/0000-0002-1739-8065 Hagens Eliza R. C. e.r.hagens@amsterdamumc.nl Feenstra Minke L. Lam Wing C. Eshuis W. J. Lameris W. van Berge Henegouwen Mark I. Gisbertz Suzanne S. s.s.gisbertz@amsterdamumc.nl grid.7177.6 0000000084992262 Department of Surgery, Cancer Center Amsterdam, Amsterdam UMC, University of Amsterdam, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands 27 4 2023 27 4 2023 2023 47 8 19952002 2 4 2023 © The Author(s) 2023 https://creativecommons.org/licenses/by/4.0/ Open AccessThis article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. Background Serum C-reactive protein (CRP) is commonly used by surgeons to raise suspicion of anastomotic leakage and other infectious complications, but most studies on optimal cut-off values are retrospective with a small sample of patients. The aim of this study was to determine the accuracy and optimal cut-off value of CRP for anastomotic leakage in patients following esophagectomy for cancer. Materials and methods Consecutive minimally invasive esophagectomy for esophageal cancer patients was included in this prospective study. Anastomotic leakage was confirmed if a defect or leakage of oral contrast was seen on a CT scan, by endoscopy or if saliva was draining from the neck incision. Diagnostic accuracy of CRP was assessed by receiver operator curve (ROC) analysis. Youden’s index was adopted to determine the cut-off value. Results A total of 200 patients were included between 2016 and 2018. Postoperative day 5 showed the highest area under the ROC (0.825) and optimal cut-off value of 120 mg/L. This resulted in a sensitivity of 75%, specificity of 82%, negative predicting value of 97%, and positive predicting value of 32%. Conclusions CRP on postoperative day 5 can be used as a negative predictor for and can be used as a marker to raise suspicion of anastomotic leakage following esophagectomy for esophageal cancer. When CRP exceeds 120 mg/L on postoperative day 5, additional investigations should be considered. issue-copyright-statement© Société Internationale de Chirurgie 2023 ==== Body pmcIntroduction Anastomotic leakage following esophagectomy for cancer is a severe complication, generally leading to additional postoperative morbidity or even death [1–4]. C-reactive protein (CRP) is a serum acute-phase reactant produced by the liver in response to pro-inflammatory cytokines that play a role in activating the complement system [5]. CRP can be tested easily at low cost and is, therefore, often used to detect infectious complications after surgery [6]. In colorectal surgery, it has been validated as a useful negative predictor test for septic complications, and cut-off values have been determined as guidance for additional diagnostic tests [7]. CRP is used by most surgeons to detect postoperative complications after esophagectomy, but the literature on optimal cut-off values are retrospective with a small sample of patients [7–10]. Identifying an optimal cut-off value for CRP for anastomotic leakage may contribute to earlier detection of a complicated postoperative course and can be used in ERAS protocols to guide the indication for additional investigations such as a CT scan or endoscopy. The primary aim of this study was to determine the accuracy and optimal cut-off values of CRP with regard to anastomotic leakage in esophageal cancer patients following minimally invasive esophagectomy. The secondary aims were to determine if there is an association between the level of CRP and the severity of anastomotic leakage and the correlation between the level of CRP and incidence of other complications. Materials and methods This prospective observational study was conducted in the Amsterdam UMC, location AMC, the Netherlands. According to national guidelines, this study was exempted from official institutional review board (IRB) approval. The IRB assessed the study design and protocol, and a waiver was obtained. Informed consent was obtained from all patients. The Standards for Reporting Diagnostic accuracy studies (STARD) guidelines were used to ensure the correct reporting of this diagnostic study [11]. Study population Consecutive patients with a histologically proven, resectable (cT0-4aN0-3M0) esophageal or gastroesophageal junction carcinoma in whom a minimally invasive esophagectomy with gastric conduit reconstruction was performed between April 2016 and October 2018 were included. Patients were excluded if a salvage esophagectomy was performed. Surgery and postoperative CRP measurements Esophagectomy was performed by means of a minimally invasive, formation of a gastric conduit, and construction of an intrathoracic or cervical anastomosis. The indication for a cervical anastomosis depended on the location of the primary tumor (≧ mid-esophageal), the clinical presence of paratracheal lymph node metastases, and the proximal extension of the radiation field (with the aim of creating the anastomosis outside the radiation field). A two-field lymphadenectomy was performed (Stations 2 on indication, 4, 5, 7, 8, 9, and 15–19 according to the AJCC 8th edition and lymph nodes in the hepatoduodenal ligament). Levels of CRP were measured on postoperative days 3, 5, and 7, according to local protocol. There may be reasons for deviating from this protocol, such as for clinical deterioration warranting determining of CRP on an earlier day or admittance to the ICU, where CRP was not routinely measured. Postoperative complications The diagnosis of anastomotic leakage was made when a defect or leakage of oral contrast was seen on a CT scan or by endoscopy or if saliva was draining from the neck incision. Final confirmation of diagnosis was made with consensus by all upper GI surgeons (SSG, MvBH, and WJE). The day of diagnosis is defined as the day anastomotic leakage was confirmed on a CT scan, by endoscopy of saliva was draining from the neck incision. Definitions and grading of anastomotic leakage and other complications were done according to the Esophageal Complications Consensus Group (ECCG) classification [12]. A type 1 leak was defined as a local defect requiring no change in therapy or treated medically or with dietary modification, a type 2 leak includes a localized defect requiring interventional but not surgical therapy, for example, interventional radiology drain, stent or bedside opening, and packing of incision, and a type 3 leak was defined as a localized defect requiring surgical therapy [12]. Outcome parameters Primary outcomes were the accuracy of CRP as a marker for anastomotic leakage and the optimal cut-off value of CRP to detect anastomotic leakage. Secondary study outcomes included the association between the severity (grade) of anastomotic leakage and the level of CRP, and the correlation between CRP and other complications. Sample size calculation The aim was to include a minimal number of events (patients who develop anastomotic leakage after surgery) of 20 in this study, since this is a generally accepted number of events for the calculation of the diagnostic accuracy of a test [13]. The incidence of anastomotic leakage in the literature varies from 3 to 25% [14, 15]. An incidence of 10% was used because the incidence of anastomotic leakage is around 10% in the study center [16]. Based on these numbers, it was determined that at least 200 patients needed to be included in this study. Statistical analysis Continuous variables with normal distributions are presented as means with standard deviations and were compared using independent t-test. Medians and interquartile ranges (IQR) were used as central tendency for continuous variables with non-normal distributions, these data were compared using the Mann–Whitney U-test. Categorical data were expressed with percentage frequencies and were compared using a Chi-square or Fisher’s exact test where appropriate. Difference in CRP levels between patients with and without complications was analyzed using the independent t-test. The diagnostic accuracy of CRP was assessed by receiver operator curve (ROC) analysis for CRP on postoperative days 3, 5, and 7. The area under the ROC (AUC) is a direct measure of the diagnostic accuracy of a test. An AUC value > 0.50 indicates the ability of a test to significantly discriminate between positive and negative cases with regard to the classification variable (e.g., presence or absence of disease). Youden’s index was adopted to determine the cut-off value in ROC analysis with highest sensitivity and specificity. Sensitivity, specificity, positive predictive value, and negative predictive value were calculated for these CRP cut-off values on postoperative days 3, 5, and 7. Kruskal–Wallis test was used to compare CRP levels in patients with different types of severity of anastomotic leakage. A P-value < 0.05 (two-sided tests) was considered significant. Missing data were handled with complete case analysis. Statistical analysis was performed with SPSS, version 24.0 for Windows (SPSS Inc., Chicago, IL). Results There were 200 patients included in this study. In 13 patients (6.5%), an open esophagectomy was performed, these patients were excluded from analyses. Baseline characteristics are shown in Table 1. With 20 patients developing anastomotic leakage, the incidence was 11%. More patients in the anastomotic leakage group had diabetes. Twenty-six percent of the patients (6 out of 23) with a cervical anastomosis developed anastomotic leakage, significantly higher than the 8.5% of the patients (14 out of 164) with anastomotic leakage with an intrathoracic anastomosis (p = 0.011). Patients with anastomotic leakage were more often readmitted within 30 days (p < 0.001) and had a higher rate of in-hospital mortality (0 patients without anastomotic leakage died in-hospital and 2 out of 20 patients with anastomotic leakage died in-hospital, p = 0.011). The postoperative day of diagnosis of anastomotic leakage ranged from 1 to 15 days, with a median of 8 days.Table 1 Baseline characteristics of all patients, and patients with and without anastomotic leakage All patients No anastomotic leakage Anastomotic leakage n = 187 n = 167 n = 20 p value Male gender 154 (82.4) 138 (82.6) 16 (80.0) 0.770 Age (years) median (IQR) 66 (59–71) 65 (59–71) 68 (62–70) 0.110 BMI (kg/m2) median (IQR) 25 (23–28) 25 (23–27) 26 (25–31) 0.055 Comorbidity  Cardiovascular 97 (51.9) 84 (50.3) 13 (65.0) 0.244  COPD 8 (4.3) 7 (4.2) 1 (5.0) 0.866  Diabetes Mellitus type 2 26 (13.9) 20 (12.0) 6 (30.0) 0.028 ASA-classification  I 53 (28.3) 48 (28.7) 5 (25.0) 0.598  II 95 (50.8) 86 (51.5) 9 (45.0)  III 39 (20.9) 33 (19.8) 6 (30.0) Histology  Adenocarcinoma 140 (74.9) 127 (76.0) 13 (65.0) 0.077  Squamous cell carcinoma 38 (20.3) 34 (20.4) 4 (20.0)  Other 9 (4.8) 6 (3.6) 3 (15.0) cT-stage  cT1 12 (6.4) 11 (6.6) 1 (5.0)  cT2 42 (22.5) 38 (22.8) 4 (20.0) 1.000  cT3 128 (68.4) 113 (67.7) 15 (75.0)  cT4 1 (0.5) 1 (0.6) 0  cTx 3 (1.6) 3 (1.8) 0  cTis 1 (0.5) 1 (0.6) 0 cN-stage  cN0 64 (34.2) 55 (32.9) 9 (45.0)  cN1 84 (44.9) 77 (46.1) 7 (35.0) 0.756  cN2 34 (18.2) 30 (18.0) 4 (20.0)  cN3 2 (1.1) 2 (1.2) 0  cNx 3 (1.6) 3 (1.8) 0 Neoadjuvant treatment  None 19 (10.2) 17 (10.2) 2 (10.0) 0.851  Chemotherapy 13 (7.0) 11 (6.6) 2 (10.0)  Chemoradiation 155 (82.9) 139 (83.2) 16 (80.0) Anastomosis  Cervical (all manual) 23 (12.3) 17 (10.2) 6 (30.0) 0.011  Intrathoracic (all stapled) 164 (87.7) 150 (89.8) 14 (70.0)  Hospital stay (days), median (IQR) 10 (8–15) 9 (8–13) 33 (18–41)  <0.001  Readmission within 30 days 28 (15.0) 19 (11.4) 9 (45.0)  <0.001  Mortality within 30 days 2 (1.1) 1 (0.6) 1 (5.0) 0.203  Hospital mortality 2 (1.1) 0 2 (10.0) 0.011 Bold values indicate statistical significance Data presented as n(%), unless indicated otherwise IQR interquartile range, BMI body mass index, COPD chronic obstructive pulmonary disease, ASA American Society of Anaesthesiologists, and MIE minimally invasive, TNM staging according to AJCC 8th edition Optimal cut-off value and accuracy of CRP to detect anastomotic leakage CRP was measured in 174 patients on day 3, in 158 patients on day 5, and in 128 patients on day 7. The mean level of CRP for patients with and without anastomotic leakage was 206 and 141 mg/L, respectively, on day 3 (p < 0.001), 174 and 91 mg/L on day 5 (p < 0.001), and 166 and 86 mg/L on day 7 (p < 0.001). Figure 1a shows the levels of CRP on postoperative days 3, 5, and 7 for patients with and without anastomotic leakage.Fig. 1 Median levels of CRP on postoperative days 3, 5, and 7 in patients with and without complications. a Anastomotic leakage and level of CRP (data displayed as mean with standard deviation). b Severity of anastomotic leakage and level of CRP (data displayed as median with interquartile range). c Postoperative complications and level of CRP (data displayed as mean with standard deviation) Table 2 shows different cut-off values for CRP on postoperative days 3, 5, and 7, and corresponding AUC, sensitivity, specificity, positive predicting value, and negative predicting value. The level of CRP on postoperative day 5 has the highest accuracy to detect anastomotic leakage, with a sensitivity of 75%, specificity of 82%, negative predicting value of 97%, and positive predicting value of 32%. The optimal cut-off value for day 5 was 120 mg/L.Table 2 CRP cut-off values and diagnostic values on POD 3, 5, and 7 for all complications and anastomotic leakage Cut-off value CRP (mg/L)a AUCROC (95% CI) P-value1 Sensitivity % (95% CI) Specificity % (95% CI) PPV% (95% CI) NPV% (95% CI) Diagnostic values for prediction of any postoperative complication in all patients CRP POD-3b CRP POD-5c CRP POD-7d 100 120 72 0.653 (0.570–0.736) 0.678 (0.595–0.761) 0.696 (0.605–0.786) 0.001  < 0.001  < 0.001 76 (68–84) 34 (24–44) 61 (51–72) 51 (39–63) 92 (86–99) 73 (58–84) 71 (63–79) 86 (75–98) 80 (70–90) 57 (44–69) 49 (40–58) 50 (38–62) Diagnostic values for prediction of anastomotic leakage in all patients CRP POD-3b CRP POD-5c CRP POD-7d 141 120 137 0.733 (0.628–0.832) 0.825 (0.735–0.916) 0.789 (0.669–0.903) 0.001  < 0.001  < 0.001 79 (61–97) 75 (54–96) 71 (48–95) 61 (53–68) 82 (76–89) 83 (76–90) 20 (11–29) 32 (17–48) 34 (17–52) 96 (92–100) 97 (94–100) 96 (92–100) Bold values indicate statistical significance CRP C-reactive protein, POD postoperative day, NVW negative predicting value, PVW positive predicting value, CI confidence interval aBased on highest Youden index, p1comparison between AUCROC and reference line, bbased on 174 patients, cbased on 158 patients, dbased on 128 patients, ebased on 154 patients, fbased on 142 patients, and gbased on 107 patients because in others CRP measurement failed on this day Severity of anastomotic leakage and level of CRP Although the median level of CRP in patients with a more severe grade of anastomotic leakage was higher, the severity of anastomotic leakage did not significantly correlate with the level of CRP (Table 3). There was also no significant difference between the level of CRP and grade of anastomotic leakage when comparing type 1 leakages with types 2 and 3 leakages together (the median levels of CRP for patients with a type 1 or type 2/3 leak were 144 and 204 mg/L, p = 0.159 on day 3; 72 and 160 mg/L, p = 0.101 on day 5; and 73 and 200 mg/L, p = 0.122 on day 7). Figure 1b shows the level of CRP for the different types of anastomotic leakage.Table 3 Median CRP level for different types of anastomotic leakage No anastomotic leakage Anastomotic leakage type 1 Anastomotic leakage type 2 Anastomotic leakage type 3 P value1 CRP POD-3 120 (15–399) 144 (79–174) 188 (101–398) 262 (188–300) 0.122 CRP POD-5 73 (15–348) 72 (69–163) 146 (86–321) 291 (160–312) 0.120 CRP POD-7 63 (11–429) 73 (38–165) 166 (65–347) 242 (200–285) 0.132 Data displayed as median (range) in mg/L. Type of anastomotic leakage according to ECCG. p1Patients without anastomotic leakage were excluded from this analysis. POD postoperative day Level of CRP and other complications One hundred and nineteen patients (64%) had at least one postoperative complication, and 36 of these patients (30%) had a complication of Clavien–Dindo grade IIIB or higher. Fifty patients (27%) had a pulmonary complication, 56 (30%) had a cardiac complication, 23 patients (12%) had chyle leakage, 6 (3%) of the patients had a urologic complication, 5 patients (3%) had vocal cord paresis, and 31 patients (17%) had another complication. The mean level of CRP for patients with and without any complication was 164 and 121 mg/L, respectively, on day 3 (p < 0.001), 118 and 71 mg/L on day 5 (p < 0.001), and 114 and 59 mg/L on day 7 (p < 0.001). Figure 1c shows the mean level of CRP for patients with any complication or no complication. Table 2 shows the optimal cut-off values for CRP on postoperative days 3, 5, and 7, and corresponding AUC, sensitivity, specificity, positive predicting value, and negative predicting value for predicting any postoperative complication. The level of CRP on postoperative day 7 had the highest AUC (0.696), with an optimal cut-off value of 72 mg/L. Discussion The primary aim of this prospective observational study was to determine the accuracy and optimal cut-off values of CRP to predict anastomotic leakage. We found that the optimal postoperative day to predict anastomotic leakage was day 5, with a cut-off value of 120 mg/L. This resulted in an AUC of 0.825, sensitivity of 75%, specificity of 82%, and a positive predictive value and negative predictive value of 32 and 97%, respectively. This is one of the few studies reporting about the correlation between CRP and anastomotic leakage after minimally invasive esophageal surgery and determining an optimal cut-off value for CRP as guidance for further diagnostic tests. The level of CRP on postoperative day 5 had the highest accuracy to detect anastomotic leakage with an optimal cut-off value of 120 mg/L. Most other studies also found postoperative day 5 the most accurate day to measure the level of CRP. Nonetheless, the cut-off value of 120 mg/L is lower than in other studies where the cut-off levels range from 154 to 189 mg/L [17–20]. Authors of these studies used similar statistical techniques to determine the optimal cut-off value. Most of these studies, however, were retrospective in design and contained a small sample of patients, with a wider variety of surgical approaches. Only one study, by Asti et al., found a lower optimal cut-off value than the present study, this was a CRP level of 83 mg/L [21]. A reason for the low cut-off value of CRP in both cohorts could be that all patients were operated by minimally invasive approach. More invasive esophageal surgery, such as an open esophagectomy, is correlated with a higher level of CRP posteroperatively [22]. The diagnostic accuracy of CRP on postoperative day 5 was high, with an AUC of 0.825. With a cut-off value of 120 mg/L, specificity and negative predicting value were 82 and 97%, respectively, while sensitivity and positive predicting value were lower (75 and 32%). Other studies also found higher specificity and negative predicting values compared to sensitivity and positive predicting values [23]. This indicates that CRP is a feasible marker to guide the use of additional investigations such as CT scanning to detect anastomotic leakage. In case of a low CRP, anastomotic leakage becomes very unlikely, and clinical observation can be continued. Although the CRP level was directly proportional with the severity of anastomotic leakage, this was not statistically significant. Due to the small number of patients in the different severity groups, there might have been insufficient power to reach significance. Another possibility could be that not the level of CRP itself, but the rapidity of rise in CRP could be an indicator for the severity of the leakage. An indicator to predict the severity of anastomotic leakage can be of clinical importance, as it could contribute to decision making in the management of leakage. Unfortunately, the number of patients was too small to look at the influence of difference of CRP level between days 3 and 5 on the severity on the leakage. Moreover, there is no available evidence to compare our results with. CRP is a marker of inflammation, and elevated levels of CRP can be caused by other inflammatory conditions, for example, pneumonia. CRP is not specific for anastomotic leakage, and other clinical symptoms should, therefore, be considered when evaluating CRP. This makes interpretation of accuracy parameters difficult to interpret. However, CRP can be used as a guide for the conditional use of postoperative CT scanning or endoscopy. In the present study, 40 patients had a CRP of 120 mg/L or higher on day 5, but only 12 of these patients (30%) had anastomotic leakage. This suggests that 69% of the patients would have received “unnecessary” additional diagnostics. However, the majority of these patients had another complication that would have been likely to find on a CT scan. CRP is an unspecific marker, feasible to detect postoperative complications, including anastomotic leakage. Four out of 16 patients with anastomotic leakage (and measurement of CRP available on day 5) did not have a CRP higher than 120 mg/L on day 5. It should be noted that in all of these patients, anastomotic leakage was detected on day 7 or later. It is possible that anastomotic leakage in these patients occurred later on than usual, and therefore, they did not have a CRP higher than 120 mg/L on day 5. Despite the high NPV of 96% in patients with a CRP < 120, additional CT scan should be performed in patients with (late) clinically suspected anastomotic leakage. A limitation of the present study was that only CRP levels of days 3, 5, and 7 were routinely measured. Levels on more postoperative days could have resulted in a cut-off value with higher diagnostic accuracy. Also, not in every patient CRP levels were available for days 3, 5, and 7 which might have caused bias. Moreover, determining only on days 3–5–7 may be insufficient for early detection of anastomotic leakage, as this may occur on the in-between days. Moreover, it would have been valuable to stratify our results for cervical and intrathoracic anastomosis, since cut-off levels might be different. Though there were only 14 patients in the group of patients with an intrathoracic anastomosis and six patients in the group of patients with a cervical anastomosis. Unfortunately, these numbers of events are too low to perform a reliable analysis on these subgroups. CRP can be a useful tool, but other symptoms such as fever can also indicate that there is an anastomotic leakage. Future studies should attempt to identify predictors for different types of anastomotic leakage and possibly stratify results for a cervical and intrathoracic anastomosis. Predicting anastomotic leakage might improve when other markers such as white blood cell count, temperature, or amylase level in drain fluid are also considered. A combination of different markers could contribute to a useful algorithm in the diagnosis and management of anastomotic leakage. Conclusion CRP on postoperative day 5 can be used as a marker to raise suspicion of anastomotic leakage in patients following an esophagectomy for esophageal cancer, but can especially be used as a negative predictor. With a cut-off value of 120 mg/L, a negative predicting value of 97% and specificity of 82% were found. When CRP exceeds 120 mg/L on postoperative day 5, additional investigations should be considered. Financial support None. Declarations Conflict of interest Van Berge Henegouwen reports to be a consultant for Johnson and Johnson, Medtronic, Stryker, and Mylan, in addition to institutional grants from Stryker and Olympus. The remaining have no conflict of interest to report. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. ==== Refs References 1. Aminian A Predictors and outcome of cervical anastomotic leakage after esophageal cancer surgery J Cancer Res Ther 2011 7 448 10.4103/0973-1482.92016 22269408 2. Van Heijl M Intrathoracic manifestations of cervical anastomotic leaks after transhiatal and transthoracic oesophagectomy Br J Surg 2010 97 726 731 10.1002/bjs.6971 20235083 3. Ye H-Y et al (2012) Personalized management of anastomotic leak after surgery for esophageal carcinoma. Chin Med Sci J Chung-kuo i hsüeh k’o hsüeh tsa chih/Chin Acad Med Sci 27:35–40. 4. Ferlay J et al. (2012) GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC CancerBase No. 11 [Internet]. Lyon, Fr. Int. Agency Res. Cancer; 2013. 10.1016/j.ucl.2013.01.011 5. Pepys MB Hirschfield GM C-reactive protein: a critical update J Clin Invest 2003 111 1805 1812 10.1172/JCI200318921 12813013 6. Adamina M Meta-analysis of the predictive value of C-reactive protein for infectious complications in abdominal surgery Br J Surg 2015 102 590 598 10.1002/bjs.9756 25776855 7. Singh PP Systematic review and meta-analysis of use of serum C-reactive protein levels to predict anastomotic leak after colorectal surgery Br J Surg 2014 101 339 346 10.1002/bjs.9354 24311257 8. Ji L Wang T Tian L Gao M The early diagnostic value of C-reactive protein for anastomotic leakage post radical gastrectomy for esophagogastric junction carcinoma: a retrospective study of 97 patients Int J Surg 2016 27 182 186 10.1016/j.ijsu.2016.02.021 26854957 9. Hagens ERC, Anderegg MCJ, van Berge Henegouwen MI, Gisbertz SS (2018) International survey on the management of anastomotic leakage after esophageal resection. Ann Thorac Surg 106:1702–1708. 10. Aiolfi A Use of C-reactive protein for the early prediction of anastomotic leak after esophagectomy: systematic review and Bayesian meta-analysis PLoS ONE 2018 13 e0209272 10.1371/journal.pone.0209272 30557392 11. Cohen JF STARD 2015 guidelines for reporting diagnostic accuracy studies: explanation and elaboration BMJ Open 2016 6 e012799 10.1136/bmjopen-2016-012799 28137831 12. Low DE International consensus on standardization of data collection for complications associated with esophagectomy Ann Surg 2015 262 286 294 10.1097/SLA.0000000000001098 25607756 13. Peduzzi P Concato J Kemper E Holford TR Feinstein AR A simulation study of the number of events per variable in logistic regression analysis J Clin Epidemiol 1996 49 1373 1379 10.1016/S0895-4356(96)00236-3 8970487 14. Blencowe NS Reporting of short-term clinical outcomes after esophagectomy a systematic review Ann Surg 2012 255 658 666 10.1097/SLA.0b013e3182480a6a 22395090 15. Wilke TJ Bhirud AR Lin C A review of the impact of preoperative chemoradiotherapy on outcome and postoperative complications in esophageal cancer patients Am J Clin Oncol 2015 38 415 421 10.1097/COC.0000000000000021 24351783 16. Slaman AE, Lagarde SM, Gisbertz SS, van Berge Henegouwen MIA (2015) Quantified scoring system for postoperative complication severity compared to the Clavien-Dindo classification the comprehensive complication index, pp 361–366. 10.1159/000433608 17. Gordon AC, Cross AJ, Foo EW, Roberts RH (2016) C-reactive protein is a useful negative predictor of anastomotic leak in oesophago-gastric resection. ANZ J Surg, pp 1–5. 10.1111/ans.13681 18. Noble F Risk assessment using a novel score to predict anastomotic leak and major complications after oesophageal resection J Gastrointest Surg 2012 16 1083 1095 10.1007/s11605-012-1867-9 22419007 19. Dutta S Fullarton GM Forshaw MJ Horgan PG McMillan DC Persistent elevation of C-reactive protein following esophagogastric cancer resection as a predictor of postoperative surgical site infectious complications World J Surg 2011 35 1017 1025 10.1007/s00268-011-1002-1 21350898 20. Prochazka V C-reactive protein as predictor of anastomotic complications after minimally invasive oesophagectomy J Minim Access Surg 2019 15 46 50 10.4103/jmas.JMAS_254_17 29595182 21. Asti E Utility of C-reactive protein as predictive biomarker of anastomotic leak after minimally invasive esophagectomy Langenbeck’s Arch Surg 2018 403 235 244 10.1007/s00423-018-1663-4 29516256 22. Koyanagi K Ozawa S Tachimori Y Minimally invasive esophagectomy in the prone position improves postoperative outcomes: role of C-reactive protein as an indicator of surgical invasiveness Esophagus 2018 15 95 102 10.1007/s10388-017-0602-8 29892934 23. Park JK Kim JJ Moon SW C-reactive protein for the early prediction of anastomotic leak after esophagectomy in both neoadjuvant and non-neoadjuvant therapy case: a propensity score matching analysis J Thorac Dis 2017 9 3693 3702 10.21037/jtd.2017.08.125 29268376