The role of lymph node dissection in kidney cancer surgery for staging and therapy

Abstract

Background Controversies on the therapeutic efficacy of lymph node dissection (LND) at time of radical nephrectomy (RN) for renal cell carcinoma (RCC) patients have not yet been solved, due to a limited knowledge of nodal dissemination patterns, thus licensing the use of unstandardized LND templates among institutions and causing post-operative patient risk-category misclassification. The aims of the thesis were I) to evaluate if the side and the location of RCC affected the probability of lymph nodal invasion (LNI) and/or nodal progression (NP) at follow-up, II) to describe nodal disease dissemination in clear cell RCC (ccRCC) patients and to assess the effect of the anatomical sites and the number of affected nodal areas on cancer specific mortality (CSM), III) to test the clinical usefulness of performing LND to stratify the risk of patients with RCC and select candidates for adjuvant treatment. Methods Core-data were represented by a prospectively collected database of 3,645 consecutive patients enrolled at a single tertiary Institution and submitted to surgery for RCC with comprehensive clinical, surgical, pathologic and follow-up data of patients. Regional LND consisted of hilar nodes plus, on the right side, pre-retro-caval nodes or, on the left side, para-aortic nodes. Extended LND consisted of regional LND plus interaortocaval nodes. Results Overall, 15% of patients underwent LND and were pN1 at surgery and during follow-up, 2.2% of patients had NP. Higher rates of LNI and NP were observed for patients with primary tumor located in more than one anatomical area relative to patients with tumor in a single area (p<0.01). Neither the RCC side nor the location reached the independent predictor status (all p>0.1). In the second study, among patients with one involved nodal site, 54 and 26% of patients were positive only in side-specific and interaortocaval station, respectively. Interaortocaval nodal positivity (HR 2.3, CI 95%: 1.3–3.9, p<0.01) represented an independent predictor of CSM. In the third study, LNI resulted as the most informative predictor of early progression (OR: 6.39; CI 95%: 3.26-12.54; p<0.0001). The accuracy was higher (p=0.008) for the model to predict early recurrence when implemented with pN (AUC: 0.76; CI95%: 0.71-0.80), as compared to the base model (AUC: 0.72; CI95%: 0.68-0.76). Patients with highrisk disease showed a large difference in the risk of progression according to pN-status (1-year risk: 58% for pN1; 31% for pN0; p<0.001). Conclusions Patients with single-side and more than one anatomical kidney area affected by RCC have higher rate of LNI at surgery and/or NP at follow-up, but this was predicted neither by side nor by location of RCC. When ccRCC patients (about 90% of all RCC) harbour nodal disease, its spreading can occur at any nodal station without involving the others. The presence of interoartocaval positive nodes does affect oncologic outcomes and therefore an extended LND template is advisable, when indicated. Moreover, performing LND at the time of RN improves risk stratification, resulting into clinical advantage for selecting high-risk patients for further treatment after surgery.