The influence of pNx/pN0 grouping in a multivariate setting for outcome modeling in patients with clear cell renal cell carcinoma

Lymphadenectomy, especially extended lymphadenectomy, is not commonly performed in patients undergoing a radical nephrectomy for clear cell renal cell carcinoma. Surgeons may sample suspicious regional lymph nodes, but the lymph node status of many patients with renal cell carcinoma remains unknown, termed stage pNx. Outcome models based on large institutional reviews have been criticized for grouping stages pNx and pN0 cases because of concern that the pNx category may include unrecognized stages pN1/pN2 disease. We evaluated cancer specific survival differences in patients with clear cell renal cell carcinoma and a lymph node stage of pNx, pN0 or pN1/pN2.

We searched the registry at our institution for patients who underwent radical nephrectomy for clear cell histology renal cell carcinoma between 1970 and 1998. Those with distant metastases at surgery were excluded from study. Clinical features obtained from the medical record included age at surgery, history of tobacco use, hypertension and symptomatic disease at presentation. A single urological pathologist reviewed all tumor specimens for nuclear grade, tumor necrosis, surgical margin status, 1997 tumor stage and lymph node status. These features were compared in patients with stages pNx and pN0 tumors. Cox proportional hazards models were used to compare cancer specific survival in univariate fashion, and after adjusting for tumor stage and grade.

The study cohort consisted of 1,535 patients with sporadic, unilateral clear cell renal cell carcinoma who underwent radical nephrectomy. There were 600 patients (39%) with stage pNx, 870 (57%) with stage pN0 and 65 (4%) with stages pN1/pN2 tumors. At an average of 4.2 years after surgery 414 patients died of renal cell carcinoma. On univariate analysis patients with stage pN0 tumors were significantly more likely to die of renal cell carcinoma than those with stage pNx tumors (risk ratio 1.40, 95% confidence interval 1.12 to 1.75, p = 0.003). However, after adjusting for tumor stage and nuclear grade the difference in outcome for stages pNx and pN0 tumors was not statistically significant (risk ratio 1.07 95% confidence interval 0.85 to 1.34, p = 0.583). Patients with stage pNx disease were significantly less likely to be symptomatic at presentation (p = 0.002), have tumors that were less than 5 cm. (p <0.001) and of lower stage (p <0.001) and grade (p = 0.005), and to have tumors with necrosis (p = 0.024) than patients with stage pN0 disease.

Combining stages pNx and pN0 cases to create outcome prediction models after radical nephrectomy for clear cell renal cell carcinoma is appropriate in a multivariate setting that includes tumor stage and grade. Clinical features available preoperatively and during surgery can help guide the decision to perform limited lymph node sampling. When the tumor is 5 cm. or greater, shows pathological necrosis or is advanced grade 3 or 4, lymph node sampling adds little prognostic information.