Multivariate Evaluation of Determinants Affecting Regional Lymph Node Metastasis and Survival in Bladder Cancer Patients Who Underwent Radical Cystectomy

Abstract

To elucidate the relative importance of clinicopathologic factors affecting regional lymph node metastasis and survival in bladder cancer patients, multivariate analyses by the logistic regression model and proportional hazards model were performed for 86 patients who underwent radical cystectomy between 1978 and 1988. Clinicopathologic factors included in the analysis were sex, age, prior tumor history, time from onset of symptoms to cystectomy, and tumor characteristics (size, number, growth pattern, grade, stage, infiltration pattern, lymphatic invasion, lymphocytic infiltration around tumors, and lymph node metastasis). Nineteen of eighty-six patients (22%) had regional lymph node metastasis. Univariate analysis showed that lymph node metastasis was related to stage (p = 0.0006), lymphatic invasion (p = 0.006) and infiltration pattern (p = 0.02). Multivariate analysis revealed that stage is the only determinant of statistical significance for lymph node metastasis. High-stage tumors (pT3a-pT4) had 15 times higher risk for lymph node metastasis than low-stage tumors (pTis-pT2). The 5-year survival rates were 25 and 75% for patients with and without lymph node metastasis, respectively. Proportional hazards model revealed that stage is the most significant determinant (p = 0.0001) for survival, followed by lymph node metastasis with borderline significance (0.005 < p < 0.1). A two-factor model consisting of stage and lymph node metastasis yielded corrected hazard ratios of 14.7 for stage and 2.3 for lymph node metastasis. The present study quantitatively confirms previous univariate analyses of factors affecting lymph node metastasis and survival in patients undergoing radical cystectomy.

Multivariate Evaluation of Prognostic Determinants for Renal Cell Carcinoma

“…The sta tistical methods used in previous studies included multi ple regression [3], logistic regression [4], and proportional hazards [5][6][7] models. Of these, the proportional hazards model has distinct advantages because follow-up data for all patients studied are available and investigators can obtain hazard ratios between unfavorable and favorable categories, these ratios indicating the relative importance of each factor quantitatively as we showed previously [ 19,20].…”

Diagnosis and treatment of pelvic lymph node metastasis in bladder cancer

“…The prognosis of the bladder cancer patients who have pelvic lymph node metastasis is poor. The 5‐year survival of these patients is reported as 7–30% 1–6 and recent works show 10‐year survival is 31–34%. 7,8 For the purpose of improving prognosis, fine needle aspiration biopsy (FNAB) on a diagnosis procedure for pelvic lymph node metastasis and multimodality therapy for positive FNAB cases were done.…”