Radiography and ultrasonography are routinely used in veterinary medicine to evaluate the kidneys. In dogs, renal size is routinely measured by radiography and ultrasonography; because of the wide variation in patient sizes, kidney lengths are compared with other anatomic measurements, rather than using the absolute measurement. Radiographically, canine kidneys are oval in shape with a homogeneous, soft tissue opacity. Renal size is evaluated radiographically by comparing the length of each renal silhouette to the length of the L2 vertebral body on a ventrodorsal view.1,2 Renal size in dogs can also be assessed ultrasonographically by comparing the length of the kidney to the aortic luminal diameter,3 or by use of CT with cortical and later medullary hyperattenuation following contrast administration.4 Glomerular filtration rate can be calculated by means of contrast-enhanced CT.5 In humans, renal size can be measured more accurately by evaluation of helical CT multiplanar reformatted images than by ultrasonography or radiography.6,7 Interpretation of renal size can be made subjectively; however, objective measurements of the kidney length are used frequently in radiography and ultrasonography, especially in teaching and aiding interpretation by students. Although availability of CT in general veterinary practice and use of CT for assessment of the abdomen in dogs have increased, objective measurements of renal size with this modality have not been established.8
The purpose of the study reported here was to determine the size of the left and right kidneys by use of CT in dogs without evidence of renal disease. We hypothesized that non–contrast-enhanced CT images could be used to assess the length of the kidneys, that these measurements would be significantly correlated with patient weight, and that measurements obtained with this method would fall within the previously published ranges for radiologic and ultrasonographic measurements.1,3,9
Materials and Methods
Case selection and medical records review
Hard copy and electronic medical records of the University of Wisconsin Veterinary Medical Teaching Hospital were searched to identify dogs that underwent CT imaging between November 1, 2007, and November 30, 2013. Dogs were included in the study if they had no evidence of abdominal disease in the medical record, with serum biochemical values reflecting kidney function within the respective laboratory reference ranges (including creatinine [range, 0.5 to 1.5 mg/dL] and BUN [range, 7 to 32 mg/dL] concentrations), and a final diagnosis of a nonrenal disease. Information collected from the medical records included breed, age, sex, and neuter status. Kidney measurements (renal length) were made for purposes of the present study by examination of stored CT images.
CT procedures
All CT examinations were performed with the dogs under general anesthesia. Drugs used for premedication and induction of anesthesia were determined on an individual basis at the discretion of the attending anesthesiologist; all dogs were intubated, and anesthesia was maintained with a halogenated ether in oxygen. During the study period, CT examinations were performed with a standard acquisition protocol through use of a single-detector row helical CT unita or a multidetector row helical CT unit.b Routine protocols included helical mode image acquisition, slice thickness of 1 to 3 mm, a scan field of view of 50 cm, and a pitch of 1.2; other settings varied according to factors such as patient size and weight. Images were acquired after the induction of short-term apnea with manual hyperventilation. Patient positioning (recumbency) varied depending on radiologist preference and clinical indication. A medium-frequency reconstruction algorithm was used.
Patient data were imported into a DICOM (Digital Imaging and Communication in Medicine) reader.c Non–contrast-enhanced dorsal oblique multiplanar reconstruction images were used to determine the longest dimension of each kidney (viewed at a window width of 350 HU and a window level of 40 HU; Figure 1), similar to methods described in human medical literature.10 Multiplanar reformatted images were produced with the computer softwarec by each individual observer to acquire a long-axis image of the kidneys for measurement of the longest dimension of each kidney. Measurements of renal length were acquired independently at 3 time points by 3 observers (2 board-certified radiologists [SJH and KRW] and a final-year veterinary medical student [BLH]). The length of the L2 vertebral body was determined on dorsal multiplanar reformatted images (viewed at a window width of 1,500 HU and a window level of 300 HU; Figure 2). The diameter of the abdominal aorta just caudal to the origin of the left renal artery was measured on transverse multiplanar reformatted images (viewed at a window width of 350 HU and a window level of 40 HU; Figure 3). The mean of the renal length values for each kidney was calculated. The ratio of mean renal length to length of L2 vertebral body and the ratio of renal length to aortic diameter were calculated.
Representative dorsal multiplanar reformatted CT image of the abdomen of a dog in a retrospective study to investigate the use of non-contrast-enhanced CT for measurement of the kidneys in 21 client-owned dogs without evidence of renal disease (window width, 350 HU; window level, 40 HU). Measurement for length of the left kidney is shown (double-headed arrow).
Citation: Journal of the American Veterinary Medical Association 248, 3; 10.2460/javma.248.3.282
Representative dorsal multiplanar reformatted CT image (window width, 1,500 HU; window level, 300 HU) depicting measurement of the length of the L2 vertebral body (double-headed arrow) in a dog.
Citation: Journal of the American Veterinary Medical Association 248, 3; 10.2460/javma.248.3.282
Representative transverse multiplanar reformatted CT image of the abdomen in a dog showing the region just caudal to the level of the renal artery (window width, 350 HU; window level, 40 HU). Measurement of aortic diameter is shown (double-headed arrow).
Citation: Journal of the American Veterinary Medical Association 248, 3; 10.2460/javma.248.3.282
Statistical analysis
Descriptive statistics (mean and median with range) were used to summarize patient demographics. Correlations of mean renal length with body weight, abdominal aorta diameter, and L2 vertebral body length were assessed by calculation of the Pearson correlation coefficient and 95% CI. Linear relationships between aortic diameter and renal length or vertebral body length and renal length were assessed with linear regression, with renal length as the response (Y variable) and aortic diameter and vertebral body length modeled as separate predictors (X variables). Intraobserver agreement for renal length of left and right kidneys was determined individually at each of the 3 time points by calculating the ICC with patient as a random effect and observer as a fixed effect (ICC [1,1]). Interobserver agreement for renal length was similarly calculated at each time point with both patient and observer used as random effects (ICC [3,1]).11 Values of ICC < 0.4 indicated poor agreement, values from 0.4 to 0.8 indicated good agreement, and values > 0.8 indicated excellent agreement. Statistical analysis was performed with publicly available software.d Values of P < 0.05 were considered significant.
Results
Twenty-one dogs met the study inclusion criteria (with measurements obtained for 42 kidneys). The study dogs were of various sizes and included 6 mixed-breed dogs, 5 Yorkshire Terriers, and 1 each of the following breeds: Doberman Pinscher, Bichon Frise, German Wirehaired Pointer, Scottish Terrier, Miniature Schnauzer, Beagle, Collie, Welsh Springer Spaniel, Labrador Retriever, and Golden Retriever. Mean age of the dogs was 4.18 years (median, 3.25 years; range, 0.42 years to 11.92 years). Twelve dogs were male (5 castrated and 7 sexually intact), and 9 were female (6 spayed and 3 sexually intact). Mean weight for the study population was 13 kg (28.6 lb; median, 9.6 kg [21.1 lb]; range, 1.9 to 36.5 kg [4.2 to 80.3 lb]).
The margins of both kidneys were determined on multiplanar reconstruction CT images for each patient, and measurements of each kidney were obtained by all observers. The mean renal length for all dogs was 5.24 cm (median, 5.11 cm; range, 2.93 to 8.56 cm).
The renal length-to-L2 vertebral body ratio was 2.7 (95% CI, 2.5 to 2.9) and fell within previously reported ranges1,9 for radiographic (2.54 to 3.45 [ventrodorsal view] and 2.38 to 3.19 [lateral view]) measurement ratios in dogs. The renal length-to-aortic diameter ratio was 7.40 (95% CI, 7.07 to 7.74) and was within the previously published range reported for ultrasonographic measurement ratios (5.5 to 9.1).3 With increase in renal length, there was a linear increase in length of the L2 vertebral body and aortic diameter (Table 1; Figure 4). There was also a strong linear association between renal length and body weight (Pearson correlation coefficient, 0.91; 95% CI, 0.80 to 0.97; P < 0.001). However, body weight did not affect the relationship between renal length and aortic diameter nor the relationship between renal length and L2 vertebral body length (P > 0.30 for both).
Pearson correlation coefficients with 95% CIs for relationships of renal length with length of the L2 vertebral body and diameter of the abdominal aorta as measured on multiplanar reformatted non-contrast-enhanced CT images of 21 dogs without renal disease.
| Variable | Correlation coefficient (r) | 95% CI | P value |
|---|---|---|---|
| Renal length-to-abdominal aorta diameter | |||
| Left kidney | 0.968 | 0.922–0.987 | < 0.001 |
| Right kidney | 0.968 | 0.921–0.987 | < 0.001 |
| Renal length-to-L2 vertebral body length | |||
| Left kidney | 0.923 | 0.816–0.969 | < 0.001 |
| Right kidney | 0.912 | 0.791–0.964 | < 0.001 |
Values of P < 0.05 were considered significant.
Scatterplots showing the relationships of abdominal aorta diameter (A) and L2 vertebral body length (B) with renal length in 21 dogs. All values represent the mean of measurements made independently on multiplanar reformatted CT images by 3 observers at 3 time points; renal values represent the mean of left and right kidneys.
Citation: Journal of the American Veterinary Medical Association 248, 3; 10.2460/javma.248.3.282
Intraobserver and interobserver agreement for renal length measurement were excellent. The intraobserver ICC ranged from 0.994 (95% CI, 0.988 to 0.997) to 0.998 (95% CI, 0.995 to 0.999). Similarly, the interobserver ICC ranged from 0.988 (95% CI, 0.971 to 0.995) to 0.993 (95% CI, 0.984 to 0.997).
Discussion
In the present study, renal length in dogs without evidence of renal disease was consistently measured by multiple reviewers on non–contrast-enhanced dorsal oblique multiplanar reformatted CT images. Radiographic kidney size is generally calculated on ventrodorsal views of the abdomen, but superimposition of bowel and border effacement can be a limitation in radiography. In previous radiographic studies in dogs,1,9 the left kidney was identified in 25 of 52 (48%) to 31 of 37 (84%) of ventrodorsal views and in 13 of 27 (48%) to 184 of 200 (92%) lateral views. Similarly, the right kidney was identified in 1 of 52 (2%) to 24 of 37 (65%) ventrodorsal views and in 2 of 27 (7%) to 180 to 200 (90%) of lateral views. In our CT study, both kidneys were identified and measured in all dogs; because measurements obtained by means of IV nephrography in dogs and in humans are larger than those measured on survey radiographs,7,9 we elected to use non–contrast-enhanced CT images. Although breathing and movement artifacts are limitations of CT,7 this was minimized in the present study because all patients were imaged under general anesthesia, and apnea was transiently induced as part of the imaging protocol.
Owing to the variability in size of our canine patients, previously described anatomic ratios used with other imaging methods1,3,9 were calculated for use in assessment of kidney size. The renal length-to-L2 vertebral body length ratio (2.7) and renal length-to-abdominal aorta diameter ratio (7.4) determined in the present study were within the previously published ranges reported for these values in radiographic1,9 and ultrasonographic3 studies involving dogs.
Gross measurements of the kidneys of dogs at necropsy in a previous study1 showed that renal weight was more highly correlated with renal length and renal volume than with renal area, body weight, body surface area, vertebral length, or vertebral area. In healthy adult people, renal parenchymal volume is strongly correlated with glomerular filtration rate.12 Similar to our findings of associations between renal length and body weight, human kidney size is influenced by body mass index, with greater body mass index associated with larger kidney size as well as by gender, with males having significantly larger kidneys than females.10,12 To our knowledge, there is no evidence of an association between sex and kidney size in dogs. In 1 study,2 the skull type of dogs was significantly associated with the renal length-to-L2 vertebral body length ratio of the left kidney, with larger values found for brachycephalic breeds, compared with dogs of other skull types.2 The population of dogs in the present study was not large enough to assess associations between renal measurements and skull type.
For measurement of renal length in human kidney donors, use of coronal plane CT images was more accurate than other methods, including analysis of standard abdominal radiographs, IV pyelography images, and ultrasonographic measurements.7 The coronal plane divides the human body into dorsal (posterior) and ventral (anterior) components, and is equivalent to the dorsal plane in quadrupeds. Comparison of human renal length measurements obtained from ultrasonographic and multiplanar reformatted CT images revealed that the ultrasonographic renal length was underestimated, compared with that obtained by CT.13 To our knowledge, this has not been investigated in dogs.
In 1 study of human patients,6 the mean absolute intraobserver and interobserver errors for renal length measurement on CT images were significantly smaller than those found for ultrasonographic images. In our study, there was excellent intraobserver and interobserver agreement for renal length on non-contrast-enhanced CT images. Renal length after contrast medium administration would likely have greater values than were determined in these non-contrast-enhanced images, in accordance with published literature. However, this comparison as well as the measurement of renal length in other planes was outside the scope of our retrospective study. Further investigation of these variables would help to confirm that findings in other species also hold true for canine patients.
Owing to the retrospective nature of this study, factors not taken into account include the effects of patient hydration or systemic arterial pressure, which may alter aortic diameter measurements. Similarly, the presence of subclinical renal disease could not be excluded in the study population. The reported ranges for renal length on radiographic and ultrasonographic images of dogs are wide and may also include some animals with undetected renal abnormalities. It should be considered that, in dogs with suspected intestinal obstruction, radiographic measurement of small intestinal diameter was not associated with increased accuracy of diagnosis, compared with subjective radiographic assessment alone,14 and it is possible that measurement of renal size or renal size ratios by CT may similarly have a low diagnostic impact. However, such measurements may prove useful, and a prospective study with a larger population of dogs, ideally including formulation of a true reference range, is needed. As CT becomes more widely available in general practice, a reference range may be useful to aid in CT assessment of the kidneys. Although not evaluated in the present study, CT can also be used to examine the kidneys for changes in size or architecture over time, and further investigation of these variables in dogs is also warranted.
Acknowledgments
Dr. Heder was a fourth-year veterinary student at the time of the study.
No third-party funding or support was received in connection with this study or the writing or publication of the manuscript. The authors declare that there were no conflicts of interest.
ABBREVIATIONS
| CI | Confidence interval |
| HU | Hounsfield unit |
| ICC | Interclass correlation coefficient |
Footnotes
GE HiSpeed LXi, GE Healthcare, Milwaukee, Wis.
GE LightSpeed 8, GE Healthcare, Milwaukee, Wis.
OsiriX 64-bit, version 3.9.4, Pixmeo, Geneva, Switzerland.
R: A language and environment for statistical computing, version 2 (2013). R Foundation for Statistical Computing, Vienna, Austria. Available at: www.r-project.org/.
References
1. Finco DR, Stiles NS, Kneller SK, et al. Radiologic estimation of kidney size of the dog. J Am Vet Med Assoc 1971; 159: 995–1002.
2. Loback MA, Sullivan M, Mellor D, et al. Effect of breed, age, weight and gender on radiographic renal size in the dog. Vet Radiol Ultrasound 2012; 53: 437–441.
3. Mareschal A, D'Anjou MA, Moreau M, et al. Ultrasonographic measurement of kidney-to-aorta ratio as a method of estimating renal size in dogs. Vet Radiol Ultrasound 2007; 48: 434–438.
4. Schwarz T. Urinary system. In: Schwarz T, Saunders J, eds. Veterinary computed tomography. Chichester, West Sussex, England: Wiley-Blackwell, 2011; 331–338.
5. Alexander K, Dunn M, Carmel E, et al. Clinical application of Patlak plot CT-GFR in animals with upper urinary tract disease. Vet Radiol Ultrasound 2011; 51: 421–427.
6. Larson D, Meyers M, O'Hara S. Reliability of renal length measurements made with ultrasound compared with measurements from helical CT multiplanar reformat images. AJR Am J Roentgenol 2011; 196: 592–597.
7. Kang K, Lee Y, Park S, et al. A comparative study of methods of estimating kidney length in kidney transplantation donors. Nephrol Dial Transplant 2007; 22: 2322–2327.
8. Fields E, Robertson I, Osborne J, et al. Comparison of abdominal computed tomography and abdominal ultrasound in sedated dogs. Vet Radiol Ultrasound 2012; 53: 513–517.
9. Lee R, Leowijuk C. Normal parameters in abdominal radiology of the dog and cat. J Small Anim Pract 1982; 23: 251–269.
10. Glodney B, Unterholzner V, Taferner B, et al. Normal kidney size and its influencing factors – a 64-slice MDCT study of 1.040 asymptomatic patients. BMC Urol 2009; 9: 19.
11. Shrout P, Fleiss J. Intraclass correlations: uses in assessing rater reliability. Psychol Bull 1979; 86: 420–428.
12. Johnson S, Rishi R, Andone A, et al. Determinants and functional significance of renal parenchymal volume in adults. Clin J Am Soc Nephrol 2011; 6: 70–76.
13. Jeffery N, Douek N, Guo D, et al. Discrepancy between radiological and pathological size of renal masses. BMC Urol 2011; 11: 2.
14. Ciasca T, David F, Lamb C. Does measurement of small intestinal diameter increase diagnostic accuracy of radiography in dogs with suspected intestinal obstruction? Vet Radiol Ultrasound 2013; 54: 207–211.
