Sentinel Lymph Node Procedure in Colon Cancer: A Systematic Review and Meta-Analysis of Individual Patient Data

Abstract

Background: The Sentinel Lymph Node (SLN) procedure has not been validated for nodal staging in colon cancer. We aimed to investigate its diagnostic performance and its effect-modifiers. Methods: We performed a systematic review and meta-analysis, searching Embase/Pubmed from database inception until March 25th, 2019 (PROSPERO CRD42016037697). Studies including ≥20 patients, showing positivity rates of SLN-procedures were eligible. We identified concept validation studies characterized by advanced histopathology of SLNs and regional lymph nodes (rLNs). We requested Individual Patient Data (IPD) to stratify for patient, tumour and procedure-related factors. Diagnostic outcome measurements were sensitivity, negative predictive value and detection rate. We also investigated the impact of procedure and tumour-related factors, and estimated the rate of upstaging and prevalence of disease in comparison with standard pathological work-up. Diagnostic outcomes were pooled using a univariate logistic random effects model, and heterogeneity was determined by I2 and Chi-square tests. Prevalence of disease was calculated as secondary outcome. Findings: We identified 47 studies with 3245 successful SLN-procedures in 3536 patients. Six high quality concept validation studies showed a sensitivity of 0.56 (95%CI:0.53-0.58) with low heterogeneity (Chi2=2.29 ; df=5 (p=0.808); I2=0.0%), negative predictive value of 0.69 (95%CI:0.64-0.73) with low heterogeneity, Chi2=10.36; df=5 (p=0.066); I2=51.7%) and detection rate of 0.85 (95%CI:0.79-0.89). Diagnostic outcomes were higher for the remaining 41 studies; sensitivity 0.74 (95%CI:0.69-0.80) with substantial heterogeneity (Chi2=119.39; df=40 (p˂0.0001); I2=66.5%), negative predictive value of 0.83 (95%CI:0.80-0.86) with substantial heterogeneity (Chi2=83.18, df=40 (p˂0.0001); I2=51.9%) and detection rate of 0.94 (95%CI:0.91-0.95). Prevalence of disease after conventional H&E-staining was 0.34 (95%CI:0.31-0.37), and 0.48 (95%CI:0.42-0.54) after advanced histopathology of all lymph nodes. Interpretation: The SLN-procedure in colon cancer is currently insufficient due to anatomical and technical difficulties, the wide variation of SLN mapping methods, patient selection and histopathology methods. Future studies should focus on low invasive tumours and real-time imaging of lymph flow towards the SLN. Upon standardization, the SLN-procedure may help to clarify the prognostic relevance of isolated tumour cells and micrometastases. Funding: None. Declaration of Interests: We declare no competing interests.