Genitourinar y Imaging Original Research

Beland et al. Renal Function in Chronic Kidney Disease Genitourinary Imaging Original Research

Renal Cortical Thickness Measured at Ultrasound: Is It Better Than Renal Length as an Indicator of Renal Function in Chronic Kidney Disease?
Michael D. Beland1 Nicholas L. Walle1 Jason T. Machan2 John J. Cronan1
Beland MD, Walle NL, Machan JT, Cronan JJ

OBJECTIVE. The purpose of our study was to determine whether there is a relationship between renal cortical thickness or length measured on ultrasound and the degree of renal impairment in chronic kidney disease (CKD). MATERIALS AND METHODS. From October to December 2007, 25 patients (13 men and 12 women, mean age 73 years) were identied who had CKD but were not on dialysis. The patients were from a single institution and had undergone renal ultrasound and at least three serum creatinines within 90 days. The lowest creatinine was used for estimated glomerular ltration rate (eGFR) calculation using both the Cockcroft-Gault (CG) and the Modication of Diet in Renal Disease Study (MDRD) equations. Ultrasounds were consensus reviewed by three radiologists (2 attendings and a resident) blinded to specic renal function. Cortical thickness was measured in the sagittal plane over a medullary pyramid, perpendicular to the capsule. Length was measured pole-to-pole. Linear regression was used for statistical analysis. RESULTS. Mean cortical thickness was 5.9 mm (range, 3.211.0 mm). Mean length was 10 cm (7.212.4 cm). Mean minimum serum creatinine was 2.1 mg/dL (1.16.1 mg/dL). Mean eGFR using CG was 34.8 mL/min (10.699.4 mL/min) and 36 mL/min (866 mL/ min) using MDRD. There was a statistically signicant relationship between eGFR and cortical thickness using both CG ( p < 0.0001) and MDRD ( p = 0.005). There was a statistically signicant relationship between CG and length ( p = 0.003) but not between MDRD and length ( p = 0.08). CONCLUSION. Cortical thickness measured on ultrasound appears to be more closely related to eGFR than renal length. Reporting cortical thickness in patients with CKD who are not on dialysis should be considered. raditional teaching is that renal length correlates with renal function in chronic kidney disease (CKD), and therefore bipolar renal lengths are almost always reported at renal ultrasound [1]. Previous studies have shown that renal volume calculated at ultrasound is a more exact measurement of a functioning kidney than renal length [2, 3]. A more recent study showed that kidney length and volume signicantly correlated with estimated glomerular ltration rate (eGFR) in the elderly, but kidney length has a low specicity in predicting renal impairment [4]. However, measuring the true kidney volume at ultrasound is difcult. Estimates of volume can be made on the basis of the ellipsoid formula, but this method has an inherent defect because the kidney is not actually ellipsoid [5]. In addition, the ellipsoid volume would include the central sinus fat

Keywords: chronic kidney disease, estimated glomerular ltration rate, renal cortex, renal failure, ultrasound DOI:10.2214/AJR.09.4104 Received December 10, 2009; accepted after revision January 21, 2010. WEB This is a Web exclusive article.
1 Department of Diagnostic Imaging, Rhode Island Hospital, Warren Alpert Medical School of Brown University, 593 Eddy St., Providence, RI 02903. Address correspondence to M. D. Beland (mbeland@lifespan.org). 2 Departments of Biostatistics and Research, Orthopaedics, and Surgery, Rhode Island Hospital, Warren Alpert Medical School of Brown University, Providence, RI.

AJR 2010; 195:W146W149 0361803X/10/1952W146 American Roentgen Ray Society

that does not contain functioning renal tissue and does vary from patient to patient. These factors likely contributed to the ndings in a recent study that showed a positive, but weak association between sonographically determined kidney volume and various indices of glomerular ltration rate (GFR) [6]. A recent study showed renal volume, and specically cortical volume, measured at CT had a strong positive relationship with renal function [7]. Additional studies have shown total renal volumes obtained at CT relate to renal function [8, 9]. However, there are drawbacks to using CT, including increased cost and radiation exposure. In addition, Widjaja et al. [8] showed a signicant correlation between ultrasound-measured renal length and CT-measured renal volume. In patients with CKD, the renal cortical echogenicity increases at ultrasound [10].

W146

AJR:195, August 2010

Renal Function in Chronic Kidney Disease In addition, the renal cortex often becomes thinned [11]. Often this nding occurs with a normal bipolar renal length and an increase in the relative amount of central sinus fat. To the best of our knowledge, a relationship between renal function and cortical thickness has not been well established at ultrasound. The purpose of our study was to determine whether there is a relationship between renal cortical thickness or length measured at ultrasound and the degree of renal impairment in CKD using two widely accepted computational methods of estimating GFR. Materials and Methods Patient Selection
This retrospective study was approved by our institutional review board and is HIPAA compliant. A search was performed of our hospital electronic medical record for patients with the clinical diagnosis of CKD who were not on dialysis and also had undergone renal ultrasound over a 3-month period from October 2007 to December 2007. This search was further narrowed to patients who had at least three serum creatinines and weight recorded within 90 days of the ultrasound. Patients with hydronephrosis were excluded. Twenty-ve patients (13 men and 12 women, mean age, 73 years; age range 2690 years) met these criteria and constitute the study population. curvilinear transducer (Logiq 9, GE Healthcare). The examinations were retrospectively reviewed at a PACS workstation (Centricity, GE Healthcare) by three authors (a radiology resident (PGY-4) and two radiology attending physicians specializing in ultrasound with 3 and 27 years of experience). All
Fig. 175-year-old woman with chronic kidney disease. Longitudinal ultrasound image of right kidney shows cortical thickness measured perpendicularly from outer margin of kidney to corticomedullary junction (arrow ). Measurement is 0.46 cm.

TABLE 1: Study Population

Patient No. Sex 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 F M M F F F M F F M M F M M F F F M M F M M M F M Minimum Maximum Mean Age Creatinine Creatinine Creatinine MDRD Mean Renal Mean Cortical (y) (mg/dL) (mg/dL) (mg/dL) eGFR CG eGFR Length (cm) Thickness (cm) 43 90 83 69 84 68 59 65 85 26 82 90 83 74 75 84 78 83 75 70 76 87 42 85 69 6.1 1.2 2.7 3.5 2.1 1.7 1.8 1.2 1.1 1.7 1.5 1.2 1.5 1.6 1.9 2.8 1.8 1.3 1.7 5.7 2.6 1.2 1.2 1.1 2.9 11.5 1.8 7.1 4.1 3.1 1.9 3.8 1.9 3.3 7.3 5.6 2.6 2.7 2.6 2.9 3.4 6.4 2.0 3.1 6.6 5.7 3.2 4.6 2.2 3.5 8.63 1.42 4.67 3.89 2.59 1.82 2.76 1.60 2.07 3.25 3.39 1.79 2.15 2.03 2.42 3.16 3.97 1.53 2.53 6.20 3.92 1.77 2.77 1.68 3.24 7.53 56.91 22.70 12.95 22.45 29.93 38.77 45.16 47.24 48.96 44.78 42.17 44.67 42.42 25.76 16.10 27.23 52.68 39.45 7.35 24.12 57.29 66.33 47.21 21.68 10.72 43.29 16.51 21.57 19.43 25.43 55.79 51.03 35.67 99.38 39.07 21.81 56.80 46.01 25.33 18.50 37.29 40.77 33.78 12.33 19.01 48.06 45.73 34.00 17.85 9.8 10.7 10.15 9.7 9.7 7.15 9.55 11.85 9.45 12.4 11.25 9.5 11.45 10.1 8.5 7.7 9.55 11.9 10.15 10.9 8.45 8.75 10.65 10.0 10.0 0.56 0.86 0.61 0.70 0.56 0.39 0.82 0.77 0.64 1.10 0.64 0.46 0.75 0.86 0.47 0.28 0.49 0.46 0.54 0.34 0.41 0.77 0.67 0.61 0.38

Estimation of Renal Function

The lowest creatinine performed within 90 days of the ultrasound was used for eGFR calculations. The lowest creatinine was chosen because it represents the best recorded renal function during the study period and helps to minimize the inuence of superimposed acute on chronic renal insufciency [10]. The Cockcroft-Gault (CG) and the Modication of Diet in Renal Disease Study (MDRD) equations were used for eGFR calculation, as follows [12]: The CG equation is eGFR = (140 age) (Weight in kg) (0.85 if female) / (72 Cr) where Cr is creatinine. The equation for Modication of Diet in Renal Disease (MDRD) for isotope dilution mass spectrometry (IDMS)-traceable creatinine measurements is GFR (mL/min/1.73 m2) = 175 (Scr) 1.154 (Age) 0.203 (0.742 if female) (1.212 if African American) (conventional units). Scr is serum creatinine.

Ultrasound Interpretation
All renal ultrasound studies were performed in the inpatient setting at our tertiary care hospital. The examinations were performed using standard gray-scale B-mode imaging with a 3.5-MHz

NoteeGFR = estimated glomerular ltration rate, MDRD eGFR = Modication of Diet in Renal Disease study for isotope dilutionmass spectometry traceable creatinine measurements, CG eGFR = Cockcroft-Gault equation. See Materials and Methods section for calculation of MDRD eGFR and CG eGFR.

AJR:195, August 2010

W147

Beland et al.
measurements were made by consensus agreement. Renal lengths were measured as the greatest poleto-pole distance in the sagittal plane. The renal cortical thickness was measured in the sagittal plane at the level of the mid kidney as described by Moghazi et al. [11]. The measurement was taken over a medullary pyramid, perpendicular to the capsule as the shortest distance from the base of the medullary pyramid to the renal capsule (Fig. 1). The readers were blinded to specic renal function, additional imaging, or any additional clinical information at the time of image review. Cortical thickness and length were measured bilaterally.
100 y = 13.5 + 80.6x, p = 0.0001 r 2 = 0.66 100 y = 46.8 + 8.2x, p = 0.0029 r 2 = 0.3

CG

50

CG
0.5

50

0 0

1.5

0 6

10

12

14

Mean Cortical Thickness (cm)

100 100 y = 9.1 + 44.7x, p = 0.0050 r 2 = 0.29

Mean Length (cm)

Statistical Analysis
The data were entered and stored on a spreadsheet (Excel, Microsoft). Mean cortical thickness and length were used in analyses. Statistical analysis and visualization were performed using Matlab (MathWorks). The relationship between ultrasound measurements and renal function was tested using linear regression. Signicance was considered at a p value < 0.05.

y = 8.7 + 4.5x, p = 0.0756 r 2 = 0.13

MDRD

50

MDRD
0.5

50

0 0

1.5

0 6

10

12

14

Results Details regarding the study population are given in Table 1. The mean cortical thickness was 5.9 mm (range, 3.211.0 mm). The mean length was 10 cm (range, 7.212.4 cm). A statistically signicant positive relationship was observed between eGFR and mean cortical thickness using both the CG and the MDRD equations (CG, p < 0.0001; MDRD, p = 0.0050). There also was a statistically signicant relationship between CG eGFR and mean renal length ( p = 0.0029) but not MDRD eGFR ( p = 0.0756) (Fig. 2). The strongest relationship, as evidenced by the highest r 2 value, was for mean cortical thickness and CG eGFR (r 2 = 66%). Discussion Our series showed a statistically signicant relationship between cortical thickness measured at ultrasound and renal function in patients with CKD. Although there was also a signicant relationship between CG eGFR and renal length, there was not for MDRD eGFR. Renal length has traditionally been considered a surrogate marker of renal function because renal length decreases with decreasing renal function. Renal lengths are universally reported and are usually the only measurements given at renal ultrasound [1]. However, on the basis of our study, it appears that cortical thickness measured at ultrasound may be related more closely to eGFR than renal length in patients with chronic renal failure.

Mean Cortical Thickness (cm)

Mean Length (cm)

Fig. 2 Graphs show Cockcroft-Gault (CG) (top) and Modication of Diet in Renal Disease (MDRD) study (bottom) estimated glomerular ltration rate (eGFR) plotted as function of mean cortical thickness (left) and mean length (right) with best-t straight line from regression.

As the burden of CKD continues to increase, efforts to reduce the cost of monitoring and managing this disease are needed. Our study attempted to evaluate the usefulness of a generally obtainable measurement at ultrasound in the setting of CKD as a correlate to kidney function (eGFR). Prior studies also have evaluated imaging measurements as surrogate markers of renal function. A study evaluating 69 patients with suspected unilateral renal artery stenosis showed renal volume was a better predictor of single-kidney GFR than renal length. They also showed the addition of renal area and parenchymal thickness measured at ultrasound to length was a better predictor of both single-kidney GFR and renal volume than length measured at ultrasound alone [8]. Another study showed a correlation between eGFR and renal volumes measured at ultrasound in 116 healthy children [13]. Other authors have described kidney volume as a better predictor of renal function than renal length [2, 3]. This was further supported by a study in 2009 by Sanusi et al. [6] showing a weak but positive correlation between kidney volume and various indices of GFR, best with measured creatinine clearance, in 40 patients with CKD. Their results also showed a signicant correlation with the measured creatinine clear-

ance and the CG and MDRD equations, further validating these estimates of GFR in CKD [6]. Our study is limited by the small study sample. We hope the results presented here will serve as a pilot study prompting further studies with larger patient samples to validate the results. Future areas of investigation using larger patient samples may include development of a predictive range of renal function given a particular cortical thickness. Alternatively, a determination of a threshold cortical thickness above which renal function is preserved may be identied. Because of the retrospective design of our study, measurements were made on the images after they were obtained. Renal cortex measurements were taken perpendicular to the renal capsule from the capsule to the corticomedullary interface. This interface can be difcult to identify in some patients in whom there is poor corticomedullary differentiation. To ensure accuracy, these measurements ideally would be made prospectively at the time of the examination. This also would allow real-time image optimization to possibly make the corticomedullary interface more apparent. Additional measurements in the transverse plane or an average cortical thickness including the upper

W148

AJR:195, August 2010

Renal Function in Chronic Kidney Disease and lower poles may prove to be a better representation of functioning parenchymal volume in future studies. Despite these limitations, we found the cortical thickness was usually easy to measure on the PACS workstation. The method used in our study also likely reects common practice in which ultrasound images are obtained by a technologist and then interpreted by the radiologist after the patient has left the department. Another potential limitation of our study is the use of computational estimates of renal function, rather than measured GFR. Although both formulas have been previously validated and are widely used, there is continued debate over which formula is best [14, 15]. A study by Rule et al. [16] in 2004 showed that the MDRD equation systematically underestimated renal function. The underestimation was much more pronounced for healthy volunteers (29% underestimation) than in patients with CKD (6.2% underestimation). However, their study was performed with a version of the MDRD equation before standardization of creatinine estimates between laboratories. The development of the IDMS-traceable MDRD study equation that was used in our study has allowed the use of standardized creatinine measurements and should minimize measurement differences [12]. A slightly more recent study directly comparing CG and MDRD estimates of GFR in patients with CKD showed the MDRD equation to be more accurate than the CG equation in patients with moderate to advanced kidney disease and diabetic nephropathy [17]. Our study only evaluated patients with CKD, which would minimize any potential underestimation of GFR using the MDRD equation relative to patients without known kidney disease and suggests MDRD may be the more accurate estimate applied to our study population [16]. Patients on dialysis were necessarily excluded from this study. Examining the relationship between renal function on the basis of serum creatinine and cortical thickness would be inherently awed in this group because the creatinine used for calculation would be a measure of dialysis efcacy rather than native renal function. Patients without known kidney disease also were not included in this study. It would be interesting to see if the correlation between cortical thickness and eGFR is also identied in these patients. Although this would be a more difcult study to perform, evaluating the applicability of measuring cortical thickness to healthy kidneys would be valuable. In summary, we have shown renal cortical thickness measured at ultrasound appears to relate to the degree of renal impairment in patients with CKD, and routine reporting of cortical thickness should be considered in such patients who are not on dialysis. References
1. American College of Radiology Website. ACR practice guideline for the performance of an ultrasound examination of the abdomen and/or retroperitoneum (in collaboration with the American Institute of Ultrasound in Medicine AIUM). www.acr.org/SecondaryMainMenuCategories/ quality_safety/guidelines/us/us_abdomen_retro. aspx. Revised 2007. Accessed April 14, 2010 2. Emamian SA, Nielson MB, Pedersen JF, Ytte L. Kidney dimensions at sonography: correlation with age, sex, and habitus in 665 adult volunteers. AJR 1993; 160:8386 3. Jones TB, Riddick LR, Harpen MD, Dubuisson RL, Samuels D. Ultrasonographic determination of renal mass and renal volume. J Ultrasound Med 1983; 2:151154 4. Van Den Noortgate N, Velghe A, Petrovic M, et al. The role of ultrasonography in the assessment of renal function in the elderly. J Nephrol 2003; 16:658662 5. Bakker J, Olree M, Kaatee R, et al. Renal volume measurements: accuracy and repeatability of US compared with that of MR imaging. Radiology 1999; 211:623628 6. Sanusi AA, Arogundade FA, Famurewa OC, et al. Relationship of ultrasonographically determined kidney volume with measured GFR, calculated creatinine clearance and other parameters in chronic kidney disease (CKD). Nephrol Dial Transplant 2009; 24:16901694 7. Muto NS, Kamishima T, Harris AA, et al. Renal cortical volume measured using automatic contouring software for computed tomography and its relationship with BMI, age and renal function. Eur J Radiol [Epub 2009 Nov 13] 8. Widjaja E, Oxtoby JW, Hale TL, Jones PW, Harden PN, McCall IW. Ultrasound measured renal length versus low dose CT volume in predicting single kidney glomerular ltration rate. Br J Radiol 2004; 77:759764 9. Herts BR, Sharma N, Lieber M, Freire M, Goldfarb DA, Poggio ED. Estimating glomerular ltration rate in kidney donors: a model constructed with renal volume measurements from donor CT scans. Radiology 2009; 252:109116 10. Khati NJ, Hill MC, Kimmel PL. The role of ultrasound in renal insufciency: the essentials. Ultrasound Q 2005; 21:227244 11. Moghazi S, Jones E, Schroepple J, et al. Correlation of renal histopathology with sonographic ndings. Kidney Int 2005; 67:15151520 12. Levey AS, Coresh J, Greene T, et al. Chronic kidney disease epidemiology collaboration. using standardized serum creatinine values in the Modication of Diet in Renal Disease study equation for estimating glomerular ltration rate. Ann Intern Med 2006; 145:247254 13. Adibi A, Adibi I, Khosravi P. Do kidney sizes in ultrasonography correlate to glomerular ltration rate in healthy children? Australas Radiol 2007; 51:555559 14. Levey AS, Coresh J, Balk E, et al. National Kidney Foundation practice guidelines for chronic kidney disease: evaluation, classication, and stratication. Ann Intern Med 2003; 139:137147 15. Stevens LA, Coresh J, Greene T, Levey AS. Assessing kidney function: measured and estimated glomerular ltration rate. N Engl J Med 2006; 354:24732483 16. Rule AD, Larson TS, Bergstralh EJ, Slezak JM, Jacobsen SJ, Cosio FG. Using serum creatinine to estimate glomerular ltration rate: accuracy in good health and in chronic kidney disease. Ann Intern Med 2004; 141:929937 17. Poggio ED, Wang X, Greene T, Van Lente F, Hall PM. Performance of the Modication of Diet in Renal Disease and Cockcroft-Gault equations in the estimation of GFR in health and in chronic kidney disease. J Am Soc Nephrol 2005; 16:459466

AJR:195, August 2010

W149