PURPOSE: To compare gadolinium-enhanced inversion-recovery magnetic resonance (MR) imaging with renal cortical scintigraphy in the diagnosis of childhood pyelonephritis. MATERIALS AND METHODS: Thirty-seven patients with fever-producing urinary tract infection underwent gadolinium-enhanced inversion-recovery MR imaging and technetium-99m renal cortical scintigraphy. Each study was read in double-blind fashion by two radiologists. The kidney was divided into three zones, and each was graded as positive, equivocal, or negative for pyelonephritis. RESULTS: Seventy kidneys (210 zones) were imaged. Twenty-six kidneys (54 zones) had evidence of pyelonephritis at both MR imaging and scintigraphy. Twenty-four kidneys (100 zones) were negative on both studies. Twelve kidneys (42 zones) were positive at MR imaging but negative at scintigraphy, and four kidneys (seven zones) were negative at MR imaging but positive at scintigraphy. The results of MR imaging for pyelonephritis were not equivalent to the results of scintigraphy (P = .001 for renal zones). The proportion of positive agreement between readers for the presence of pyelonephritis was 0.85 and 0.57 for MR imaging and scintigraphy, respectively. The proportion of negative agreement was 0.88 and 0.80 for MR imaging and scintigraphy, respectively. CONCLUSION: Gadolinium-enhanced inversion-recovery MR imaging enabled detection of more pyelonephritic lesions than did renal cortical scintigraphy and had superior interobserver agreement.
PURPOSE: To compare gadolinium-enhanced inversion-recovery magnetic resonance (MR) imaging with renal cortical scintigraphy in the diagnosis of childhood pyelonephritis. MATERIALS AND METHODS: Thirty-seven patients with fever-producing urinary tract infection underwent gadolinium-enhanced inversion-recovery MR imaging and technetium-99m renal cortical scintigraphy. Each study was read in double-blind fashion by two radiologists. The kidney was divided into three zones, and each was graded as positive, equivocal, or negative for pyelonephritis. RESULTS: Seventy kidneys (210 zones) were imaged. Twenty-six kidneys (54 zones) had evidence of pyelonephritis at both MR imaging and scintigraphy. Twenty-four kidneys (100 zones) were negative on both studies. Twelve kidneys (42 zones) were positive at MR imaging but negative at scintigraphy, and four kidneys (seven zones) were negative at MR imaging but positive at scintigraphy. The results of MR imaging for pyelonephritis were not equivalent to the results of scintigraphy (P = .001 for renal zones). The proportion of positive agreement between readers for the presence of pyelonephritis was 0.85 and 0.57 for MR imaging and scintigraphy, respectively. The proportion of negative agreement was 0.88 and 0.80 for MR imaging and scintigraphy, respectively. CONCLUSION:Gadolinium-enhanced inversion-recovery MR imaging enabled detection of more pyelonephritic lesions than did renal cortical scintigraphy and had superior interobserver agreement.
Authors: J Damien Grattan-Smith; Marcos R Perez-Bayfield; Richard A Jones; Stephen Little; Bruce Broecker; Edwin A Smith; Hal C Scherz; Andrew J Kirsch Journal: Pediatr Radiol Date: 2003-03-13