- Email: [email protected]
Can Hounsfield unit values of the cortex and papillae determined by computed tomography demonstrate the possibility of kidney stone formation?
European Journal of Radiology 81 (2012) 1446–1449
Contents lists available at ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
Can Hounsfield unit values of the cortex and papillae determined by computed tomography demonstrate the possibility of kidney stone formation? I˙ smet Baran a , Nuray Voyvoda b,∗ , Özlem Tokgöz a , Hüsnü Tokgöz c a
Zonguldak Karaelmas University School of Medicine, Department of Radiology, Turkey Kocaeli Acıbadem Hospital, Department of Radiology, Turkey c Zonguldak Karaelmas University School of Medicine, Department of Urology, Turkey b
a r t i c l e
i n f o
Article history: Received 30 December 2010 Accepted 15 March 2011 Keywords: Kidney calculi Computed tomography Kidney papilla Pathogenesis
a b s t r a c t Purpose: This study is aimed at measuring HU values of the renal cortex and papillae in patients with nephrolithiasis and demonstrating renal changes associated with nephrolithiasis. Materials and methods: Measurements were performed with regard to HU values of the cortex and papillae of 82 patients with unilateral nephrolithiasis and 81 patients in the control group at the level of the upper pole, middle region and lower pole of both kidneys. Results: When the HU values obtained from the upper pole, middle region and lower pole of the kidney with calculi and unaffected kidney in patients with nephrolithiasis were compared with those for the control group, the difference among the groups were found to be significant (p < 0.001). A comparison of the cortex and papillae densities of the affected and unaffected kidneys in patients with unilateral nephrolithiasis were compared with regard to the upper pole and middle region, no statistically significant difference was observed with regard to both the cortex and papillae densities of the upper pole, middle region. However, in those patients with calculi in the lower pole, the region with calculi has a higher papillae density as compared to the unaffected region. Conclusion: Both kidneys in patients with calculi have a comparatively high renal cortex and papillae densities. In the future, this information may be useful in predicting which patients may develop nephrolithiasis. © 2011 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Renal colic is a common presenting symptom in hospital emergency departments. The lifetime incidence of nephrolithiasis is estimated at 12%, and the cost of this disease in the United States is estimated at $1.83 billion [1]. Calcium oxalate is the commonest type of calculus. The pathogenesis of renal calculus, according to Randall’s theory, arises from calcium phosphate deposits lying immediately under the papillary epithelium which initiate the adherence of more calcium and eventually develop into renal stone [2]. Sites of crystal deposition at or near the renal papillae are known as Randall’s plaque and occur in 100% of patients with calcium oxalate stones formers [3]. The proportion of papillary surface coverage by Randall’s plaque also correlates with the number of stones formed among the stone forming group and significantly less in the non-stone former, when examined with
∗ Corresponding author. Tel.: +90 505 2365068. E-mail addresses: [email protected] (I˙ . Baran), [email protected] (N. Voyvoda), [email protected] (Ö. Tokgöz), h [email protected] (H. Tokgöz). 0720-048X/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2011.03.054
nephroscopic papillary mapping [4]. Randall’s plaque theory has been investigated extensively with biopsies either with percutaneous nephrolithotomy or endoscopically [5]. However, these methods are invasive. It is argued that recurrence can be determined by measuring the Hounsfield unit of the renal papillae by means of CT, which is a non-invasive method [6]. CT has emerged as the modality of choice in the evaluation of acute flank pain. It is safe, rapid, and accurate, with one series reporting 96% sensitivity, 99% specificity, and 98% accuracy in the diagnosis of urolithiasis [7]. The clinical management and follow up of patients who present with single renal calculus varies according to their risk factors and clinical symptoms. Some patients who present with single episode of renal colic do not have long-term follow up by the urologist. However there will be a group of patients among this subgroup that may be predisposed to recurrent stone formations, requiring long-term follow up and treatment to prevent repeat attacks [6]. This study is aimed at measuring HU values of the renal cortex and papillae in patients with nephrolithiasis and demonstrating renal changes associated with nephrolithiasis.
I˙ . Baran et al. / European Journal of Radiology 81 (2012) 1446–1449
1447
Fig. 1. Measurement of the HU values of the cortex and papillae for the control group at the level of the upper pole, the middle region and the lower pole.
2. Materials and methods 2.1. Subjects
middle and lower pole renal cortex and papillae were measured separately in both the kidneys in patient with nephrolithiasis and control subjects (Fig. 1).
Institutional review board approval was obtained prior to initiation of the study. A total of 348 patients were evaluated who were admitted to our hospital with renal colic complaints and subjected to CT scanning. 88 cases with bilateral nephrolithiasis, 35 who were administered contrast media orally or intravenously, 33 whose creatinine values were above 1.6 mg/dl, 32 with a catheter in the urinary tract, 24 having calculi in the ureter, 16 with many cysts in the kidneys, 14 suffering from renal fusion anomaly, 13 with nephrectomy, 6 with solitary kidneys, 3 having calculi in the urinary bladder and 2 with renal hypoplasia were all excluded from the study. 82 patients with unilateral nephrolithiasis were included in the study. The control group covered 81 cases with normal renal function in which no calculus or anomaly was detected in the urinary system on CT.
The statistical evaluation was conducted by means of SPSS (version 13.0). How well numerical variables fit a normal distribution was visualized using the Kolmogorov–Smirov test. The definitive statistics for numerical variables were expressed in average ± standard deviation and those for categorical data in numbers. For analysis of the differences between the groups with regard to variables determined by measurement, the Mann–Whitney Utest and independent samples t-test were used. In comparing the dependent variables determined by measurement, the Wilcoxon signed-rank test was employed. The variations in the measurement variables were investigated by means of correlation analysis. A 95% confidence interval was obtained when evaluating the results. Results were considered significant when p < 0.05.
2.2. CT protocol
3. Results
The CT scans were obtained by means of a Philips Secure Spiral CT scanner and a Toshiba Activion 16 CT scanner. No oral or intravenous contrast was administered. The scans were performed by using a standard stone viewing protocol at 120–140 kV, 79–260 mA, a section in the 3–5 mm range and a section thickness of 3–5 mm. The scan length included from the level of diaphragm to the pubic symphysis.
Our study included a total of 82 cases with unilateral nephrolithiasis, 46 men and 36 women. The patients were in the age range of 17–86 (mean 47). 34 of the patients had calculi located in the right renal calyx system (5 in the upper calyx, 17 in the middle calyx and 12 in the lower calyx) and 48 cases in the left renal calyx system (7 in the upper calyx, 17 in the middle calyx and 24 in the lower calyx). A total of 12 cases had calculi located in the upper calyx system, 34 in the middle calyx system and 36 in the lower calyx system. The creatinine values of the patients ranged from 0.7 mg/dl to 1.4 mg/dl. 18 patients had hypertension and 10 diabetes mellitus. The control group consisted of a total of 81 cases, 53 of whom were men and 28 women. They were in the age range of 16–86 and had an average age of 51. The creatinine values for the control group were in the 0.4–1.4 mg/dl range. 26 cases suffered from hypertension, 9 diabetes mellitus, 4 coronary artery disease and 3 hyperlipidemia. Patients with nephrolithiasis and control patients were similar with respect to mean age and baseline creatinine levels. The HU values of the cortex and papillae of the kidney with calculi and those of the kidney without calculi in patients with renal calculi were compared for the upper, middle and lower calyxes. No
2.3. Image analysis The CT images were retrospectively reviewed by two radiologists (I˙ .B, NV) in same reading session. In evaluating patients with unilateral calculi, first their localization and size were determined. Coronal reformatted images were used for better definition of the papillary anatomy and the density measurements were obtained on coronal images. The images were magnified to 5× to prevent contamination of the region of interest with the fat in the renal sinus. Renal papilla and cortex Hounsfield densities were measured by placing region of interests (mean size – 0.2 cm2 ) in the region of renal papilla and cortex. Attenuation measurements were recorded (Hounsfield Unit – HU values). The HU values of the one upper,
2.4. Statistical analysis
I˙ . Baran et al. / European Journal of Radiology 81 (2012) 1446–1449
1448
Table 1 Comparison of the HU values of the cortex and papillae of the right kidney and unaffected left kidney in the cases with right nephrolithiasis with the HU values for the control group. Hasta (n:34) RK cortex upper RK cortex middle RK cortex lower RK papillae upper RK papillae middle RK papillae lower LK cortex upper LK cortex middle LK cortex lower LK papillae upper LK papillae middle LK papillae lower
29.71 33.74 33.38 35.97 41.74 41.76 29.40 32.28 31.98 36.55 40.32 40.30
± ± ± ± ± ± ± ± ± ± ± ±
4.21 6.69 7.30 7.04 7.68 8.20 3.62 6.17 7.15 5.91 6.65 7.00
Control (n:81) 26.12 26.94 26.86 26.91 26.44 27.10 25.60 26.25 26.75 26.49 26.41 26.53
± ± ± ± ± ± ± ± ± ± ± ±
2.76 2.53 3.03 3.31 2.66 2.91 2.71 2.82 2.57 2.70 2.44 2.81
p <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001
RK: right kidney, LK: left kidney.
significant difference was observed between the HU values for the cortex and papillae in both the upper and lower calyxes (p > 0.05). A comparison of the HU values for the cortex and papillae of the patients with calculi in the lower calyx of the right kidney with those for the lower region of the left kidney revealed a significant difference (p values for the cortex and papillae are 0.032 and 0.003, respectively). When the values of the cortex and papillae in the patients with calculi in the lower calyx of the left kidney were compared with those for the lower region of the right kidney, no significant difference was observed with regard to the HU values for the cortex, whereas the HU values for the papillae were found to be significantly different (p = 0.016). When the HU values for the cortex and papillae obtained from the upper pole, the middle region and lower pole of both the right kidney and the unaffected left kidney in the cases with right nephrolithiasis were compared with those for the control group, all the measurements revealed that the difference between the two groups was significant (p < 0.001) (Table 1). A comparison of the HU values of the cortex and papillae obtained from the upper pole, middle region and the lower pole of the affected left kidney and right kidney free of calculi in the cases with left nephrolithiasis with those of the control group found that, on the basis of all the measurements, there existed a significant difference between the two groups (p < 0.001) (Table 2). When a HU cutoff value of 31 was determined for cortex density as Bhuskute et al. did, the positive predictive value (ppv) was found to be 73%, the negative predictive value (npv) 70%, the sensitivity 69% and the specificity 74% [6]. When 34 was determined to be HU cutoff value based on the same study, the positive predictive value, negative predictive value, sensitivity and speci-
Table 2 Comparison of the HU values of the cortex and papillae of the left kidney and the unaffected right kidney in the cases with left nephrolithiasis with the HU values for the control group. Hasta (n:48) LK cortex upper LK cortex middle LK cortex lower LK papillae upper LK papillae middle LK papillae lower RK cortex upper RK cortex middle RK cortex lower RK papillae upper RK papillae middle RK papillae lower
29.58 31.83 31.67 37.17 39.88 40.88 29.75 31.65 30.90 35.92 38.77 39.38
LK: left kidney, RK: right kidney.
± ± ± ± ± ± ± ± ± ± ± ±
3.95 5.82 6.69 5.13 6.61 7.32 3.25 5.42 4.20 6.69 8.02 7.06
Control (n:81) 25.60 26.25 26.75 26.49 26.41 26.53 26.12 26.94 26.86 26.91 26.44 27.10
± ± ± ± ± ± ± ± ± ± ± ±
2.71 2.82 2.57 2.70 2.44 2.81 2.76 2.53 3.03 3.31 2.66 2.91
p <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001
ficity were established to be 93%, 79%, 89% and 93%, respectively [6]. 4. Discussion Seventy years ago, Randall examined the papillae of cadaveric renal units and demonstrated that interstitial crystal plaques in the papillary tip were common in stone formers [8]. Other studies on Randall’s plaques identified crystalline deposits in the renal papillae of patients with nephrolithiasis [9–12]. Furthermore, endoscopic mapping studies have associated with the severity of calcium nephrolithiasis with the area of plaque coverage, an increase in plaque area results in larger calculi and increased number of stone events [2,11]. The present study investigates the effect of calculi on the renal cortex and papillae using a non-invasive radiological method on the basis of numeric data such as Hounsfield unit values. In renal pathology, density measurement on the basis of the Hounsfield unit is valuable in distinguishing the renal calculi from blood clots and gaining insight into the composition of renal calculi [13,14]. The literature contains little work on HU values of the renal papillae in patients with and without nephrolithiasis [5,6]. The studies Eisner et al. conducted on 17 patients and a control group found that the papillae densities of both kidneys in the cases with nephrolithiasis were higher than that of the control group. The control group was established to have more concomitant diseases compared to the patient group with nephrolithiasis, a factor that might affect the results of their studies. In our study, co morbidity among the patient and control groups was kept concordant. In a similar study carried out by Bhuskute et al. on 90 patients and a control group, it was established that both the affected and unaffected kidneys in the nephrolithiasis patients had a higher papillae density than those in the control group. The cutoff values for the cortex and papillae which they employed in their study were found to have low ppv (19% and 33%) and high npv (98% and 99%). In our study in which we used the cutoff values for the cortex and papillae determined by Bhuskute et al., we achieved a higher ppv (73% and 93%) and a lower npv (70% and 79%). We sought to demonstrate that, owing to a relatively higher ppv, papillae density in particular could be used as a scanning test for renal calculi. The facts that a comparison of HU values of the cortex and papillae for the patients with calculi in the lower pole of the right kidney with those for the lower region of the left kidney found a significant difference (p values for the cortex and papillae were 0.032 and 0.003, respectively), and a similar comparison made for those patients with calculi in the lower region of the left kidney revealed a significant difference with regard to the papillae but not the cortex, although magnification was performed in HU measurement, are attributable to the fact that the partial volume effect was not completely eliminated. In our study, a correlation was established between the measurements of the cortex and papillae. Therefore, it is thought that, in measuring HU values with respect to regions where calculi occur, only cortex measurements will be enough to reduce partial volume effect. There are no differences between the affected and non-affected sides within the stone-formers group, which implies that the process is systemic and it affects both kidneys symmetrically. 5. Conclusion This study has shown that radiologists potentially can provide clinicians with the risk indicator of stone formation in
I˙ . Baran et al. / European Journal of Radiology 81 (2012) 1446–1449
a routinely performed CT examination for renal colic. Appearance of HU values for renal parenchyma in reports will be useful in following up with high risk patients. In addition, a prospective study is required accompanied by knowledge of stone subtypes, patient dehydration status, urine volume and urinary calcium. References [1] Clark JY, Thompson IM, Optenberg SA. Economic impact of urolithiasis in the United States. J Urol 1995;154:2020–4. [2] Evan AP, Lingeman JE, Coe FL, et al. Randall’s plaque of patients with nephrolithiasis begins in basement membranes of thin loops of Henle. J Clin Invest 2003;111(5):607–16. [3] Low RK, Stoller ML. Endoscopic mapping of renal papillae for Randall’s plaques in patients with urinary stone disease. J Urol 1997;158(6):2062–4. [4] Kim SC, Coe FL, Tinmouth WW, et al. Stone formation is proportional to papillary surface coverage by Randall’s plaque. J Urol 2005;173(1):117–9. [5] Eisner BH, Iqbal A, Namasivayam S, et al. Differences in computed tomography density of the renal papillae of stone formers and non-stone formers: a pilot study. J Endourol 2008;22(10):2207–10.
1449
[6] Bhuskute NM, Yap WW, Wah TM. A retrospective evaluation of Randall’s plaque theory of nephrolithiasis with CT attenuation values. Eur J Radiol 2009;72:470–2. [7] Dalrymple NC, Verga M, Anderson KR, et al. The value of unenhanced helical computerized tomography in the management of acute flank pain. J Urol 1998;159:735–40. [8] Randall A. The origin and growth of renal calculi. Ann Surg 1937;105:1009–27. [9] Williams JC, Matlaga BR, Kim SC, et al. Calcium oxalate calculi found attached to the renal papilla: preliminary evidence for early mechanisms in stone formation. J Endourol 2006;20:885. [10] Evan A, Lingeman J, Coe FL, Worcester E. Randall’s plaque: pathogenesis and role in calcium oxalate nephrolithiasis. Kidney Int 2006;69:1313. [11] Matlaga BR, Coe FL, Evan AP, Lingeman JE. The role of Randall’s plaques in the pathogenesis of calcium stones. J Urol 2007:177–231. [12] Evan AE, Lingeman JE, Coe FL, et al. Histopathology and surgical anatomy of patients with primary hyperparathyroidism and calcium phosphate stones. Kidney Int 2008;4:223. [13] Garant M, Bonaldi VM, Taourel P, Pinsky MF, Bret PM. Enhancement patterns of renal masses during multiphase helical CT acquisitions. Abdom Imaging 1998;23:431. [14] Demirel A, Suma S. The efficacy of non-contrast helical computed tomography in the prediction of urinary stone composition in vivo. J Int Med Res 2003;31:1–5.
Contents lists available at ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
Can Hounsfield unit values of the cortex and papillae determined by computed tomography demonstrate the possibility of kidney stone formation? I˙ smet Baran a , Nuray Voyvoda b,∗ , Özlem Tokgöz a , Hüsnü Tokgöz c a
Zonguldak Karaelmas University School of Medicine, Department of Radiology, Turkey Kocaeli Acıbadem Hospital, Department of Radiology, Turkey c Zonguldak Karaelmas University School of Medicine, Department of Urology, Turkey b
a r t i c l e
i n f o
Article history: Received 30 December 2010 Accepted 15 March 2011 Keywords: Kidney calculi Computed tomography Kidney papilla Pathogenesis
a b s t r a c t Purpose: This study is aimed at measuring HU values of the renal cortex and papillae in patients with nephrolithiasis and demonstrating renal changes associated with nephrolithiasis. Materials and methods: Measurements were performed with regard to HU values of the cortex and papillae of 82 patients with unilateral nephrolithiasis and 81 patients in the control group at the level of the upper pole, middle region and lower pole of both kidneys. Results: When the HU values obtained from the upper pole, middle region and lower pole of the kidney with calculi and unaffected kidney in patients with nephrolithiasis were compared with those for the control group, the difference among the groups were found to be significant (p < 0.001). A comparison of the cortex and papillae densities of the affected and unaffected kidneys in patients with unilateral nephrolithiasis were compared with regard to the upper pole and middle region, no statistically significant difference was observed with regard to both the cortex and papillae densities of the upper pole, middle region. However, in those patients with calculi in the lower pole, the region with calculi has a higher papillae density as compared to the unaffected region. Conclusion: Both kidneys in patients with calculi have a comparatively high renal cortex and papillae densities. In the future, this information may be useful in predicting which patients may develop nephrolithiasis. © 2011 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Renal colic is a common presenting symptom in hospital emergency departments. The lifetime incidence of nephrolithiasis is estimated at 12%, and the cost of this disease in the United States is estimated at $1.83 billion [1]. Calcium oxalate is the commonest type of calculus. The pathogenesis of renal calculus, according to Randall’s theory, arises from calcium phosphate deposits lying immediately under the papillary epithelium which initiate the adherence of more calcium and eventually develop into renal stone [2]. Sites of crystal deposition at or near the renal papillae are known as Randall’s plaque and occur in 100% of patients with calcium oxalate stones formers [3]. The proportion of papillary surface coverage by Randall’s plaque also correlates with the number of stones formed among the stone forming group and significantly less in the non-stone former, when examined with
∗ Corresponding author. Tel.: +90 505 2365068. E-mail addresses: [email protected] (I˙ . Baran), [email protected] (N. Voyvoda), [email protected] (Ö. Tokgöz), h [email protected] (H. Tokgöz). 0720-048X/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2011.03.054
nephroscopic papillary mapping [4]. Randall’s plaque theory has been investigated extensively with biopsies either with percutaneous nephrolithotomy or endoscopically [5]. However, these methods are invasive. It is argued that recurrence can be determined by measuring the Hounsfield unit of the renal papillae by means of CT, which is a non-invasive method [6]. CT has emerged as the modality of choice in the evaluation of acute flank pain. It is safe, rapid, and accurate, with one series reporting 96% sensitivity, 99% specificity, and 98% accuracy in the diagnosis of urolithiasis [7]. The clinical management and follow up of patients who present with single renal calculus varies according to their risk factors and clinical symptoms. Some patients who present with single episode of renal colic do not have long-term follow up by the urologist. However there will be a group of patients among this subgroup that may be predisposed to recurrent stone formations, requiring long-term follow up and treatment to prevent repeat attacks [6]. This study is aimed at measuring HU values of the renal cortex and papillae in patients with nephrolithiasis and demonstrating renal changes associated with nephrolithiasis.
I˙ . Baran et al. / European Journal of Radiology 81 (2012) 1446–1449
1447
Fig. 1. Measurement of the HU values of the cortex and papillae for the control group at the level of the upper pole, the middle region and the lower pole.
2. Materials and methods 2.1. Subjects
middle and lower pole renal cortex and papillae were measured separately in both the kidneys in patient with nephrolithiasis and control subjects (Fig. 1).
Institutional review board approval was obtained prior to initiation of the study. A total of 348 patients were evaluated who were admitted to our hospital with renal colic complaints and subjected to CT scanning. 88 cases with bilateral nephrolithiasis, 35 who were administered contrast media orally or intravenously, 33 whose creatinine values were above 1.6 mg/dl, 32 with a catheter in the urinary tract, 24 having calculi in the ureter, 16 with many cysts in the kidneys, 14 suffering from renal fusion anomaly, 13 with nephrectomy, 6 with solitary kidneys, 3 having calculi in the urinary bladder and 2 with renal hypoplasia were all excluded from the study. 82 patients with unilateral nephrolithiasis were included in the study. The control group covered 81 cases with normal renal function in which no calculus or anomaly was detected in the urinary system on CT.
The statistical evaluation was conducted by means of SPSS (version 13.0). How well numerical variables fit a normal distribution was visualized using the Kolmogorov–Smirov test. The definitive statistics for numerical variables were expressed in average ± standard deviation and those for categorical data in numbers. For analysis of the differences between the groups with regard to variables determined by measurement, the Mann–Whitney Utest and independent samples t-test were used. In comparing the dependent variables determined by measurement, the Wilcoxon signed-rank test was employed. The variations in the measurement variables were investigated by means of correlation analysis. A 95% confidence interval was obtained when evaluating the results. Results were considered significant when p < 0.05.
2.2. CT protocol
3. Results
The CT scans were obtained by means of a Philips Secure Spiral CT scanner and a Toshiba Activion 16 CT scanner. No oral or intravenous contrast was administered. The scans were performed by using a standard stone viewing protocol at 120–140 kV, 79–260 mA, a section in the 3–5 mm range and a section thickness of 3–5 mm. The scan length included from the level of diaphragm to the pubic symphysis.
Our study included a total of 82 cases with unilateral nephrolithiasis, 46 men and 36 women. The patients were in the age range of 17–86 (mean 47). 34 of the patients had calculi located in the right renal calyx system (5 in the upper calyx, 17 in the middle calyx and 12 in the lower calyx) and 48 cases in the left renal calyx system (7 in the upper calyx, 17 in the middle calyx and 24 in the lower calyx). A total of 12 cases had calculi located in the upper calyx system, 34 in the middle calyx system and 36 in the lower calyx system. The creatinine values of the patients ranged from 0.7 mg/dl to 1.4 mg/dl. 18 patients had hypertension and 10 diabetes mellitus. The control group consisted of a total of 81 cases, 53 of whom were men and 28 women. They were in the age range of 16–86 and had an average age of 51. The creatinine values for the control group were in the 0.4–1.4 mg/dl range. 26 cases suffered from hypertension, 9 diabetes mellitus, 4 coronary artery disease and 3 hyperlipidemia. Patients with nephrolithiasis and control patients were similar with respect to mean age and baseline creatinine levels. The HU values of the cortex and papillae of the kidney with calculi and those of the kidney without calculi in patients with renal calculi were compared for the upper, middle and lower calyxes. No
2.3. Image analysis The CT images were retrospectively reviewed by two radiologists (I˙ .B, NV) in same reading session. In evaluating patients with unilateral calculi, first their localization and size were determined. Coronal reformatted images were used for better definition of the papillary anatomy and the density measurements were obtained on coronal images. The images were magnified to 5× to prevent contamination of the region of interest with the fat in the renal sinus. Renal papilla and cortex Hounsfield densities were measured by placing region of interests (mean size – 0.2 cm2 ) in the region of renal papilla and cortex. Attenuation measurements were recorded (Hounsfield Unit – HU values). The HU values of the one upper,
2.4. Statistical analysis
I˙ . Baran et al. / European Journal of Radiology 81 (2012) 1446–1449
1448
Table 1 Comparison of the HU values of the cortex and papillae of the right kidney and unaffected left kidney in the cases with right nephrolithiasis with the HU values for the control group. Hasta (n:34) RK cortex upper RK cortex middle RK cortex lower RK papillae upper RK papillae middle RK papillae lower LK cortex upper LK cortex middle LK cortex lower LK papillae upper LK papillae middle LK papillae lower
29.71 33.74 33.38 35.97 41.74 41.76 29.40 32.28 31.98 36.55 40.32 40.30
± ± ± ± ± ± ± ± ± ± ± ±
4.21 6.69 7.30 7.04 7.68 8.20 3.62 6.17 7.15 5.91 6.65 7.00
Control (n:81) 26.12 26.94 26.86 26.91 26.44 27.10 25.60 26.25 26.75 26.49 26.41 26.53
± ± ± ± ± ± ± ± ± ± ± ±
2.76 2.53 3.03 3.31 2.66 2.91 2.71 2.82 2.57 2.70 2.44 2.81
p <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001
RK: right kidney, LK: left kidney.
significant difference was observed between the HU values for the cortex and papillae in both the upper and lower calyxes (p > 0.05). A comparison of the HU values for the cortex and papillae of the patients with calculi in the lower calyx of the right kidney with those for the lower region of the left kidney revealed a significant difference (p values for the cortex and papillae are 0.032 and 0.003, respectively). When the values of the cortex and papillae in the patients with calculi in the lower calyx of the left kidney were compared with those for the lower region of the right kidney, no significant difference was observed with regard to the HU values for the cortex, whereas the HU values for the papillae were found to be significantly different (p = 0.016). When the HU values for the cortex and papillae obtained from the upper pole, the middle region and lower pole of both the right kidney and the unaffected left kidney in the cases with right nephrolithiasis were compared with those for the control group, all the measurements revealed that the difference between the two groups was significant (p < 0.001) (Table 1). A comparison of the HU values of the cortex and papillae obtained from the upper pole, middle region and the lower pole of the affected left kidney and right kidney free of calculi in the cases with left nephrolithiasis with those of the control group found that, on the basis of all the measurements, there existed a significant difference between the two groups (p < 0.001) (Table 2). When a HU cutoff value of 31 was determined for cortex density as Bhuskute et al. did, the positive predictive value (ppv) was found to be 73%, the negative predictive value (npv) 70%, the sensitivity 69% and the specificity 74% [6]. When 34 was determined to be HU cutoff value based on the same study, the positive predictive value, negative predictive value, sensitivity and speci-
Table 2 Comparison of the HU values of the cortex and papillae of the left kidney and the unaffected right kidney in the cases with left nephrolithiasis with the HU values for the control group. Hasta (n:48) LK cortex upper LK cortex middle LK cortex lower LK papillae upper LK papillae middle LK papillae lower RK cortex upper RK cortex middle RK cortex lower RK papillae upper RK papillae middle RK papillae lower
29.58 31.83 31.67 37.17 39.88 40.88 29.75 31.65 30.90 35.92 38.77 39.38
LK: left kidney, RK: right kidney.
± ± ± ± ± ± ± ± ± ± ± ±
3.95 5.82 6.69 5.13 6.61 7.32 3.25 5.42 4.20 6.69 8.02 7.06
Control (n:81) 25.60 26.25 26.75 26.49 26.41 26.53 26.12 26.94 26.86 26.91 26.44 27.10
± ± ± ± ± ± ± ± ± ± ± ±
2.71 2.82 2.57 2.70 2.44 2.81 2.76 2.53 3.03 3.31 2.66 2.91
p <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001
ficity were established to be 93%, 79%, 89% and 93%, respectively [6]. 4. Discussion Seventy years ago, Randall examined the papillae of cadaveric renal units and demonstrated that interstitial crystal plaques in the papillary tip were common in stone formers [8]. Other studies on Randall’s plaques identified crystalline deposits in the renal papillae of patients with nephrolithiasis [9–12]. Furthermore, endoscopic mapping studies have associated with the severity of calcium nephrolithiasis with the area of plaque coverage, an increase in plaque area results in larger calculi and increased number of stone events [2,11]. The present study investigates the effect of calculi on the renal cortex and papillae using a non-invasive radiological method on the basis of numeric data such as Hounsfield unit values. In renal pathology, density measurement on the basis of the Hounsfield unit is valuable in distinguishing the renal calculi from blood clots and gaining insight into the composition of renal calculi [13,14]. The literature contains little work on HU values of the renal papillae in patients with and without nephrolithiasis [5,6]. The studies Eisner et al. conducted on 17 patients and a control group found that the papillae densities of both kidneys in the cases with nephrolithiasis were higher than that of the control group. The control group was established to have more concomitant diseases compared to the patient group with nephrolithiasis, a factor that might affect the results of their studies. In our study, co morbidity among the patient and control groups was kept concordant. In a similar study carried out by Bhuskute et al. on 90 patients and a control group, it was established that both the affected and unaffected kidneys in the nephrolithiasis patients had a higher papillae density than those in the control group. The cutoff values for the cortex and papillae which they employed in their study were found to have low ppv (19% and 33%) and high npv (98% and 99%). In our study in which we used the cutoff values for the cortex and papillae determined by Bhuskute et al., we achieved a higher ppv (73% and 93%) and a lower npv (70% and 79%). We sought to demonstrate that, owing to a relatively higher ppv, papillae density in particular could be used as a scanning test for renal calculi. The facts that a comparison of HU values of the cortex and papillae for the patients with calculi in the lower pole of the right kidney with those for the lower region of the left kidney found a significant difference (p values for the cortex and papillae were 0.032 and 0.003, respectively), and a similar comparison made for those patients with calculi in the lower region of the left kidney revealed a significant difference with regard to the papillae but not the cortex, although magnification was performed in HU measurement, are attributable to the fact that the partial volume effect was not completely eliminated. In our study, a correlation was established between the measurements of the cortex and papillae. Therefore, it is thought that, in measuring HU values with respect to regions where calculi occur, only cortex measurements will be enough to reduce partial volume effect. There are no differences between the affected and non-affected sides within the stone-formers group, which implies that the process is systemic and it affects both kidneys symmetrically. 5. Conclusion This study has shown that radiologists potentially can provide clinicians with the risk indicator of stone formation in
I˙ . Baran et al. / European Journal of Radiology 81 (2012) 1446–1449
a routinely performed CT examination for renal colic. Appearance of HU values for renal parenchyma in reports will be useful in following up with high risk patients. In addition, a prospective study is required accompanied by knowledge of stone subtypes, patient dehydration status, urine volume and urinary calcium. References [1] Clark JY, Thompson IM, Optenberg SA. Economic impact of urolithiasis in the United States. J Urol 1995;154:2020–4. [2] Evan AP, Lingeman JE, Coe FL, et al. Randall’s plaque of patients with nephrolithiasis begins in basement membranes of thin loops of Henle. J Clin Invest 2003;111(5):607–16. [3] Low RK, Stoller ML. Endoscopic mapping of renal papillae for Randall’s plaques in patients with urinary stone disease. J Urol 1997;158(6):2062–4. [4] Kim SC, Coe FL, Tinmouth WW, et al. Stone formation is proportional to papillary surface coverage by Randall’s plaque. J Urol 2005;173(1):117–9. [5] Eisner BH, Iqbal A, Namasivayam S, et al. Differences in computed tomography density of the renal papillae of stone formers and non-stone formers: a pilot study. J Endourol 2008;22(10):2207–10.
1449
[6] Bhuskute NM, Yap WW, Wah TM. A retrospective evaluation of Randall’s plaque theory of nephrolithiasis with CT attenuation values. Eur J Radiol 2009;72:470–2. [7] Dalrymple NC, Verga M, Anderson KR, et al. The value of unenhanced helical computerized tomography in the management of acute flank pain. J Urol 1998;159:735–40. [8] Randall A. The origin and growth of renal calculi. Ann Surg 1937;105:1009–27. [9] Williams JC, Matlaga BR, Kim SC, et al. Calcium oxalate calculi found attached to the renal papilla: preliminary evidence for early mechanisms in stone formation. J Endourol 2006;20:885. [10] Evan A, Lingeman J, Coe FL, Worcester E. Randall’s plaque: pathogenesis and role in calcium oxalate nephrolithiasis. Kidney Int 2006;69:1313. [11] Matlaga BR, Coe FL, Evan AP, Lingeman JE. The role of Randall’s plaques in the pathogenesis of calcium stones. J Urol 2007:177–231. [12] Evan AE, Lingeman JE, Coe FL, et al. Histopathology and surgical anatomy of patients with primary hyperparathyroidism and calcium phosphate stones. Kidney Int 2008;4:223. [13] Garant M, Bonaldi VM, Taourel P, Pinsky MF, Bret PM. Enhancement patterns of renal masses during multiphase helical CT acquisitions. Abdom Imaging 1998;23:431. [14] Demirel A, Suma S. The efficacy of non-contrast helical computed tomography in the prediction of urinary stone composition in vivo. J Int Med Res 2003;31:1–5.