Abstract
C-reactive protein may be overexpressed in kidney cancer tissue. We determined if urine C-reactive protein concentration is greater in patients with high grade renal cell carcinoma or greater than healthy controls, and compared to other urologic cancers and non-cancerous urologic disease. This observational study obtained pre-nephrectomy urine and plasma from 87 patients with a small imaged renal mass (≤4cm) undergoing partial nephrectomy. Postoperative pathology established clear cell or papillary renal cell cancer, angiomyolipoma, chromophobe, or oncocytoma, and tumor Fuhrman grade. Follow-up urine was collected from 52 clear cell/papillary cancer patients. Urine and/or plasma were from bladder or prostate cancer patients and matched healthy controls. C-reactive protein was measured by enzyme-linked immunosorbent assay. Urine C-reactive protein in grade 1-2 (n=41) and grade 3-4 (n=24) renal cancer was 17- and 80-fold greater than controls, respectively. This decreased 85% post-nephrectomy, suggesting origin from tumors and not systemic filtration from blood. Urine C-reactive protein in angiomyolipoma, chromophobe, or oncocytoma, bladder or prostate cancer was less than grade 3-4 renal cancer. C-statistic to differentiate grade 1-2 from 3-4 clear cell/papillary renal cancer was 0.98 (95% CI 0.95-1.00); sensitivity 1.00 (95% CI, 0.87-1.00), specificity 0.95 (95% CI 0.84-0.99). Two patients with remarkably high pre-nephrectomy urine C-reactive protein, but no others, developed late metastases. Pre-nephrectomy urine C-reactive protein was increased in patients with high grade clear cell and papillary renal cancer, but not other renal or urologic disease. Urine C-reactive protein may be used pre-surgically to noninvasively grade small renal masses, and may predict future renal cell cancer metastasis.