The Diagnostic Value of a Quick MRI for Renal Scars and Infection
Emily L. Davidson, MD, Ruthie Su, MD, MS, Kara Gill, MD, Steve Cho, MD, Ellen Wald, MD.
University of Wisconsin-Madison, Madison, WI, USA.
BACKGROUND – Technetium 99m-labeled dimercaptosuccinic acid (DMSA) is the current standard imaging test to evaluate for renal scars but requires sedation, IV access, time, and involves ionizing radiation. We developed an abbreviated unenhanced MRI protocol (“quick MRI (qMRI)”) to screen for renal scars and pyelonephritis. The test requires <15 minutes and avoids IV access and radiation. We evaluated the accuracy of qMRI in identifying renal scars or pyelonephritis using the DMSA scan as the reference standard. METHODS – Patients scheduled for DMSA to screen for renal scars were recruited to undergo qMRI. All qMRIs were performed without sedation. Patients hospitalized with pyelonephritis who had DMSA and qMRI were also included. One radiologist read the DMSA and one radiologist read the qMRI. Both were blinded to the clinical history. Each kidney was divided into 12 segments and each segment was scored according to a 5 point Likert scale (5=Certain normal, 4=Probably normal, 3=Indeterminate, 2=Probably abnormal, 1=Certain abnormal). The overall quality of the study was also scored (5=Excellent, 4=Good, 3=Moderate, sufficient for diagnosis, 2=Poor, diagnostic confidence substantially reduced, 1= Bad, no diagnosis possible). We calculated the sensitivity, specificity, and accuracy with 95% confidence intervals after excluding indeterminate scores. RESULTS – For 6 of the 8 cases, DMSA and MRI were completed within 7 days of each other. Six patients underwent screening for renal scar; scars were detected on DMSA in three patients. Of 132 kidney zones, 8% percent were scored as indeterminate on DMSA compared to 18% on qMRI. The median study quality score for the DMSA scans was 5 (“excellent”) and 3 (“Moderate, sufficient for diagnosis”) for the MRI scans. The sensitivity, specificity, and accuracy of qMRI for scar was 31% [9%-61%], 97% [92%-99%], and 89% [81%-94%] respectively- and after excluding 2 cases with poor and bad quality- 80% [28%-99%],97% [90%-99%], and 96% [89%-99%] respectively. Two patients with clinical pyelonephritis had both studies during their hospital stay. The image quality for both DMSA and qMRI was good or excellent. Of 48 kidney zones, 47 zones were scored. The cortical defects were identified as scars on DMSA but on qMRI they were consistent with pyelonephritis and in one patient, a focal intrarenal abscess was found. The sensitivity, specificity, and accuracy of MRI in detecting pyelonephritis was 88% [47%-100%], 87% [73%-96%], and 87% [74%-95%] respectively. CONCLUSIONS –The qMRI was highly specific in detecting renal scars; mild sedation of young children may improve immobility and therefore its diagnostic accuracy for renal scars. The qMRI was accurate in detecting pyelonephritis and allowed for objective differentiation of pyelonephritis from scar. Further experience and protocol refinement will make qMRI a potential alternative to DMSA to detect and localize renal cortical abnormalities.
Back to 2023 Abstracts