Eur. Radiol.

8, 212217 (1998) Springer-Verlag 1998

European Radiology

Original article Renal colic: comparison of spiral CT, US and IVU in the detection of ureteral calculi
S. Yilmaz, T. Sindel, G. Arslan, C. Ozkaynak, K. Karaali, A. Kabaalioglu, E. Luleci
Department of Radiology, Akdeniz University Medical Faculty, Arapsuyu, TR-07 070 Antalya, Turkey Received 20 May 1997; Revision received 18 July 1997; Accepted 22 July 1997

Abstract. The aim of our study was to compare noncontrast spiral CT, US and intravenous urography (IVU) in the evaluation of patients with renal colic for the diagnosis of ureteral calculi. During a period of 17 months, 112 patients with renal colic were examined with spiral CT, US and IVU. Fifteen patients were lost to follow-up and excluded. The remaining 97 patients were defined to be either true positive or negative for ureterolithiasis based on the follow-up data. Sensitivity, specificity, positive and negative predictive value and accuracy of spiral CT, US and IVU were determined, and secondary signs of ureteral stones and other pathologies causing renal colic detected with these modalities were noted. Of 97 patients, 64 were confirmed to have ureteral calculi based on stone recovery or urological interventions. Thirty-three patients were proved not to have ureteral calculi based on failure to recover a stone and diagnoses unrelated to ureterolithiasis. Spiral CTwas found to be the best modality for depicting ureteral stones with a sensitivity of 94 % and a specificity of 97 %. For US and IVU, these figures were 19, 97, 52, and 94 %, respectively. Spiral CT is superior to US and IVU in the demonstration of ureteral calculi in patients with renal colic, but because of its high cost, higher radiation dose and high workload, it should be reserved for cases where US and IVU do not show the cause of symptoms. Key words: Ureter calculi urography CT stenosis or obstruction

inexpensive and repeatable technique which allows direct demonstration of urinary stones located at ureterovesical junction (UVJ), and in the renal pelvis or calices. However, stones located between the renal pelvis and UVJ are extremely difficult to demonstrate with US, and very small stones may not be seen because of its technical limitations. The IVU technique has the advantages of showing the stones located at any level in the urinary system, and giving physiological information on the degree of obstruction. However, IVU may not show some ureteral stones because of their low attenuation value or bowel superimposition [3, 4]. Recently, several articles have shown that spiral CT is very useful for depicting ureteral stones and other causes of renal colic [1, 2, 5, 6]. Comparative studies were previously performed between spiral CT vs IVU, and spiral CT vs a combination of US and plain films in a limited number of patients [1, 2]. This study was intended to compare the value of spiral CT, US and IVU for the demonstration of ureteral stones in a relatively large group of individuals. Materials and methods From June 1995 to December 1996, 112 patients with symptoms of renal colic were examined on the same day with noncontrast CT, US and IVU. Informed consent was obtained from each patient prior to inclusion in the study. Fifteen patients were excluded since they were unable to be contacted for follow-up; the remaining 97 patients (54 males and 43 females; mean age 41.2 years) constituted our study group. These patients were then followed-up in an attempt to obtain the final diagnosis, and the course of clinical symptoms, passage of a calculus, results of urological interventions and diagnoses other than ureterolithiasis were noted. All spiral CT examinations were performed with an Xpress CT scanner (Toshiba, Tokyo, Japan). No oral or IV contrast material was used. Images were obtained from the top of the kidneys to the bladder base, using

Introduction When a patient presents with renal colic, US and intravenous urography (IVU) are the most commonly used diagnostic modalities [1, 2]. Ultrasound is a safe, rapid,
Correspondence to: S. Yilmaz

S. Yilmaz et al.: Renal colic

213

two to three clusters of 21-s duration at 200 mAs and 120 kV. Table feed rate was 5 mm per rotation, and nominal section thickness was 5 mm (pitch = 1). Image reconstruction was routinely performed at 5-mm increments by means of 360 interpolation. When a calcific density was seen along the course of the ureter, further evaluation of this region was performed using retrospective reconstructions in 2-mm increments. In some patients reformatted images were also obtained using overlapped sections with 2-mm intervals, to create images similar to IVU. All spiral CT images were interpreted by two radiologists (S. Y. and G. A.). Ultrasonography was performed with a colour Doppler machine (Toshiba SSA 270 A, Tokyo, Japan) using a 3.5-MHz convex probe. Ultrasonographic examination included evaluation of the renal pelvis and kidneys, followed by ureterovesical junctions (UVJ) through a full urinary bladder. Absent or asymmetrical ureteral jets were also noted for both sides. All US examinations were performed and interpreted by the same radiologist (AK). For IVU, patients received oral purgative the night before. After a plain abdominal film was obtained, 50 ml contrast media was injected intravenously, and films were taken at 2nd, 5th and 15th minutes. When a delay in calyceal or ureteral opacification was present, late films up to 24 hours were also obtained. IVU images were interpreted by two radiologists (T. S. and K. K.). Our criteria for the diagnosis of ureterolithiasis on noncontrast helical CT included (a) a calcific density which appeared to be within the ureter lumen, and visualization of proximal ureter to its level, and/or presence of a tissue rim sign [2], and (b) a calcific density located at the anatomic site of ureterovesical junction (UVJ). The US diagnosis of a ureteric stone required visualization of an echogenic focus within ureteropelvic junction (UPJ), proximal ureter or UVJ. Diagnosis of a ureteric stone on IVU is made if an opacity is seen along the course of the ureter and the proximal ureter can be traced to the level of the opacity, and/or there is unilateral opacification delay. The presence or absence of secondary signs of ureteral stones on spiral CT, US and IVU were also noted. They included calyceal and ureteral dilatation, perinephric stranding, perinephric fluid, increase in renal size, residual renal calculi, absent or asymmetric ureteral jets (on colour Doppler) and unilateral opacification delay (on IVU). Secondary signs were considered to be present only if they were located on the symptomatic side. However, unless a ureteric stone was demonstrated, diagnosis of ureterolithiasis was not made on the basis of these secondary signs, since none of them was specific for ureterolithiasis. After the imaging procedures were completed, patients were followed-up in order to confirm or exclude the diagnosis of ureterolithiasis. Based on the follow-up data, patients in whom the passage of a calculus was documented, or a ureteric stone was recovered by urological interventions, were considered to be true positive for ureterolithiasis. Those who failed to recover a

stone despite the disappearance of clinical symptoms, and those in whom pathologies unrelated to ureterolithiasis were detected, were considered to be true negative. In both true-positive and true-negative patients, presence or absence of direct and supportive findings as well as unrelated pathologies were noted by interpreters who were blinded to clinical symptoms and the results of other imaging modalities. For spiral CT and IVU, images were first interpreted separately and then simultaneously by two radiologists, with final decisions made by consensus. For direct demonstration of ureteral stones, sensitivity, specificity, positive and negative predictive value and accuracy of spiral CT, US and IVU were calculated. Results Of 97 patients, 64 were determined to have ureteral calculi. In 42 patients, clinical symptoms disappeared following the passage of a calculus, and the control images showed the suspected stones to be absent in their previous locations. In 22 patients, presence of ureteral stones was confirmed by percutaneous lithotripsy (n = 13) or ureteroscopic stone extraction (n = 9). The remaining 33 patients were determined not to have ureteral calculi. None of these patients recovered a stone. In 7 patients, pathology unrelated to urinary stone disease was demonstrated; 3 patients had appendicitis, 2 adnexal masses, 1 common bile duct stone and 1 diverticulitis. Three patients were treated successfully for urinary tract infection, which was thought to account for their symptoms. In the remaining 23 patients, the reason for the flank pain could not be demonstrated. In all these patients, clinical symptoms disappeared within 1 week. In 64 patients with proven ureteral stones, spiral CT directly demonstrated ureteral stones in 60, US in 12 and IVU in 33 (Figs. 14). In 33 patients without ureteral stones, spiral CT was falsely positive in 1 patient, US in 1 and IVU in 2. In the cases where spiral CT and US was falsely positive, control CT and US did not show the suspected stones in their previous locations. These stones measured 23 mm in size and were thought to have passed but were missed by the patients. In 2 cases where IVU was falsely positive, spiral CT showed no stones within the ureter, but it showed phleboliths in the suspected region. These opacities were thought to be phleboliths superimposed on the ureter. For direct

Table 1. Sensitivity, specificity, accuracy and positive and negative predictive values of spiral CT, US and IVU in the direct demonstration of ureteral stones % Sensitivity 94 19 52 Specificity 97 97 94 Accuracy Positive predictive value 98 92 94 Negative predictive value 89 38 50

Spiral CT US IVU

95 45 66

214

S. Yilmaz et al.: Renal colic

1a

1b

2a

2b

2c

2d
with proved ureteral stones but no stone seen on CT, there were 13 secondary signs. Pathologies that mimicked ureteral stone disease were detected more frequently with spiral CT; in 3 cases with appendicitis, spiral CT showed the lesion in all, and US in 1, 2 cases with adnexal mass were successfully diagnosed with both spiral CT and US, and in 1 case with a common bile duct stone, the lesion could be seen only with spiral CT. Intravenous urography was normal in all these cases. In 1 patient with diverticulitis, none of these modalities showed any abnormality (Table 3). All these diagnoses were confirmed by laparoscopy or surgery.
Table 2. Detection rates of secondary signs of ureteral stones with spiral CT, US and IVU in patients with stones Spiral CT Ureteral dilatation Calyceal dilatation Stranding of perinephric fat tissue Perinephric fluid Renal stones Nephromegaly Delay in ureteral opacification Absent or asymmetric ureteral jets 54 (84) 50 (78) 38 (70) 8 (12) 16 (25) 34 (53) US 11 (17) 47 (73) 0 6 (9) 5 (8) 29 (45) IVU 48 (75) 44 (69) 0 0 7 (11) 18 (28) 48 (75) 34 (53)

Fig. 1 a, b. Direct demonstration of a stone within the ureter and its secondary signs on noncontrast spiral CT. a A stone within the left ureter with a prominent tissue rim sign. b An upper section shows dilatation of proximal ureter and a residual renal stone Fig. 2ad. Detection of nonopaque calculi with spiral CT in a patient with right renal colic. a Plain abdominal film shows no urinary stones. b Spiral CT images show a renal stone. c A ureteral stone with a tissue rim sign. d Reformatted spiral CT image clearly demonstrates both calculi and slight dilatation of the collecting system. In this patient, renal US only showed slight calicial dilatation and decreased ureteral jets

demonstration of ureteral stones, spiral CT proved to be the best modality with 94 % sensitivity, 97 % specificity, 95 % accuracy, 98 % positive predictive value and 89 % negative predictive value (Table 1). Detection rates of secondary signs of ureteral stones by spiral CT, US and IVU in patients with stones are given in Table 2. According to these figures, ureteral and calyceal dilatation, perinephric stranding and fluid, nephromegaly and residual renal stones were detected more frequently with spiral CT than with US and IVU. Unilateral delay in ureteral opacification and absent or asymmetric ureteral jets could be evaluated only with IVU and colour Doppler US, respectively. In 5 patients of 60 with proved ureteral stones and stones seen on CT, none of the secondary signs was present on spiral CT, US and IVU, whereas in 55, at least one secondary sign could be seen on spiral CT. In 5 patients with no secondary signs, spiral CT showed a calculus in all, IVU in 2 and US in 1 (Fig. 3). In 4 patients

NOTE: Numbers in parentheses are percentages

S. Yilmaz et al.: Renal colic

215

3a

3b

4a

3c

4b
Fig. 3ac. A patient with right renal colic. a Plain abdominal film. b IVU show no ureteral stones or secondary signs. c Spiral CT image shows a ureteral stone located at the right ureterovesical junction, which was recovered by the patient the following day. This stone could not be seen with US either

4c

4d
Discussion

Fig. 4ad. A patient with left renal colic. a Plain abdominal film shows two opacities, superimposed on the left kidney and ureter. b Spiral CT image shows the renal stone and a large parapelvic cyst with wall calcification. c, d Reformatted spiral CT images clearly demonstrate the renal stone, cyst calcification and ureteral calculus, as well as another ureteral stone located proximally, which is not seen on the plain abdominal film Table 3. Detection frequencies of pathologies unrelated to ureterolithiasis by spiral CT, US and IVU Pathologies Appendicitis (n = 3) Adnexal mass (n = 2) Common bile duct stone (n = 1) Diverticulitis (n = 1) Spiral CT 3 2 1 0 US 1 2 0 0 IVU 0 0 0 0

Several recent studies have demonstrated that noncontrast spiral CT is the preferred imaging technique for determining the presence or absence of ureteral calculi in patients with suspected renal colic [1, 2, 5, 6]. In their pioneering study, Smith et al. [1] compared noncontrast CT and IVU in 20 patients, and found noncontrast CT to be more effective than IVU in identifiying ureteral stones, and equally effective as IVU in the determination of presence or absence of ureteral obstruction. They also reported that after the completion of their study, clinicians became reluctant to refer such patients for IVU examination. In another comparative study which included 34 patients, Sommer et al. [2] reported that reformatted noncontrast spiral CT images were superior to the combination of US and plain abdominal film in the demonstration of ureteral calculi. In our study, a similar comparison was made among spiral CT, US and IVU in a group of 97 patients, and in accordance with these previous reports, spiral CT proved to be the

216

S. Yilmaz et al.: Renal colic

best modality in the demonstration of ureteral stones with 94 % sensitivity, 97 % specificity, 98 % positive predictive value and 89 % negative predictive value (Table 1). In the study by Smith et al. [5], these figures were 97, 96, 96 and 97 %, respectively. The results of both studies show that spiral CT is extremely useful in the demonstration as well as exclusion of ureterolithiasis. In our study, the specificity and positive predictive values of US and IVU were comparable to that of spiral CT, which suggests that if a ureteral stone is seen with US or IVU, then spiral CT is probably unnecessary to confirm the diagnosis. However, both US and IVU had a low sensitivity and negative predictive value compared with that of spiral CT, indicating that they are not ideal for the exclusion of ureterolithiasis. Thus, if US and IVU are normal, a ureteral stone cannot be excluded unless spiral CT is performed. The low sensitivity and negative predictive value of US can be explained by the fact that the part of ureters between ureteropelvic junction and UVJ are very difficult to visualize with this technique. Similarly, plain films may not show ureteral stones in many cases, either due to bowel superimposition or because their contrast resolution is not sufficient to delineate stones with a low calcium content. Spiral CT has the advantages of both providing axial images free from superimposition, and having a high contrast resolution which results in the demonstration of any kind of urinary stones with higher attenuation values (200600 HU) than surrounding soft tissue [3, 7, 8]. Our criteria for the diagnosis of ureterolithiasis were based on the direct demonstration of a stone within the lumen of the ureter. Direct demonstration of ureteral stones offers some advantages. First of all, if a stone is seen to be located within the ureter, diagnosis of ureterolithiasis is straightforward. In these patients, analgesic treatment and hydration can be started without any need for further diagnostic procedures. Secondly, the size of the stone can be determined, and the probability of spontaneous passage can be estimated; approximately 80 % of all stones smaller than 5 mm will pass spontaneously, but this rate diminishes as the stone size increases [9]. Thirdly, the level of the ureteral stone can be determined so that the type of intervention could be decided. In our study, one of the criteria used for intraureteral location of a stone was the presence of a tissue rim sign, which is seen as a ring of soft tissue density surrounding a ureteral stone (Fig. 1 a). This finding is thought to represent the local edema of the ureteral wall, and was found to be 92100 % specific for showing the intraureteral location of a calculus [10, 11]. In our study, tissue rim sign was present in 49 of 60 patients with ureteral stones seen on CT, and in all, the patients were proved to have ureteral calculi. Secondary signs of ureteral stones may be helpful when diagnosis of ureterolithiasis is uncertain (Fig. 1 b). However, they mostly indicate ureteral obstruction and are by no means specific for ureteral stones. As reported previously, pathologies other than ureterolithiasis may cause ureteral obstruction and mimic renal colic. Value

of the secondary signs that can be seen on CT are reported by Smith et al. [6] who found that unilateral ureteral dilatation was the most useful finding in the diagnosis of ureterolithiasis with a sensitivity of 90 % and a specificity of 93 %. In our study, 5 patients of 60 with proved stones showed none of the secondary findings, although spiral CT clearly showed a calculus in all cases (Fig. 3). In this group, IVU showed a stone in 2 cases, and US in 1. It is obvious that if only secondary signs had been used as diagnostic criteria, 3 of these stones would have been missed with IVU and 4 with US. Sommer et al. [2] reported to have three similar patients with ureteral stones, in whom US examination was normal because of the lack of significant ureteral obstruction and hydronephrosis. Smith et al. [5, 6] reported that if no ureteral stone is demonstrated and both ureteral dilatation and stranding are present on the symptomatic side in a patient, the recent passage of a calculus or a calculus that cannot be seen should be considered. In our study, this combination of secondary signs was present in 2 of 4 patients who recovered a stone (all 23 mm in size) but had no stones seen on CT. Several factors may be responsible for the nonvisualization of stones in these cases, including volume averaging, small stone size and low attenuation value of the stones. In our study, calyceal and ureteral dilatation, perinephric stranding, perinephric fluid, increase in renal size and residual renal calculi were more frequently detected with spiral CT (Fig. 1 b). The difference is particularly pronounced in the demonstration of perinephric stranding and residual renal calculi (Table 2). In both instances, the superior performance of CT can be attributed to its increased contrast resolution. Absent or asymmetric ureteral jets can be evaluated only with colour Doppler US, and unilateral opacification delay can be determined only with IVU, unless IV contrast is given during spiral CT. In our study, no further statistics was performed as to the detection rates of secondary signs with spiral CT, US and IVU because of the lack of an external gold standard, and of the fact that some of the secondary findings could not be evaluated by all the modalities. Extraurinary causes mimicking ureterolithiasis were also more frequently demonstrated with spiral CT in our study. However, the small number of such cases precluded any statistical analysis of these results. Although the value of spiral CT in the demonstration of these pathologies is well known, it is not possible to conclude that spiral CT is superior to US and IVU in this respect, based solely on our results. In our study, the only case that spiral CT overlooked was diverticulitis in a 46year-old man, which was diagnosed at surgery. In their series, Smith et al. [5] successfully diagnosed a number of other extraurinary pathologies with spiral CT, such as liver hemangioma and leaking abdominal aortic aneurysm, as well as abnormalities of the urinary tract unrelated to ureterolithiasis such as pyelonephritis, emphysematous cystitis and renal malignancies. The main limitation of our study is the fact that in most cases, determination of true-negative and true-

S. Yilmaz et al.: Renal colic

217

positive cases was based on the observation of the patients as to whether they recovered a stone or not. This applied for 42 of 64 true-positive and 26 of 33 true-negative patients. If we consider that some small stones could have been missed by the patients, then the number of false-negative cases will have been underestimated, and that of false-positive cases overestimated. In approximately half of our patients, reformatted spiral CT images were also created (Figs. 2 d and 4 c, d). As indicated by Sommer et al. [2], the quality of these images depend greatly on the degree of ureteral dilatation and the amount of retroperitoneal fat. In their study, Sommer et al. [2] used curved planar reformatting technique which demonstrated the full course of ureters. In our study, because of our software limitation, reformatted images were obtained using basic multiplanar reconstruction technique, creating images only in straight planes. For this reason, reformatted images were mainly utilized to obtain images similar to IVU, which facilitated communication with clinicans, and were not routinely used for diagnostic purposes. Spiral CT has some distinct advantages in the evaluation of patients with renal colic. As indicated by previous studies, it is the most valuable imaging modality for depicting ureteral stones and for the demonstration of their secondary signs, as well as other pathologies mimicking urinary stone disease. It does not require oral purgative, and IV or oral contrast material which may interfere with the performance of other diagnostic procedures. Thus, it can safely be used in patients with intolerance to contrast media or oral purgatives. If an abnormality is questionable, the examination can be repeated with oral or IV contrast, or IVU can be performed. And finally, examination time of spiral CT is approximately 5 min, even in patients with obstruction, which is slightly shorter than that of US, whereas a routine IVU requires at least 20 min, or sometimes much longer depending on the number of late films taken. The main disadvantage of spiral CT is its inability to give information on renal function. However, once a ureteral stone is seen on spiral CT, IVU or spiral CT with contrast can be performed to show decreased or delayed opacification. Radiation dose of spiral CT is generally higher than that of IVU. At our institution,

skin dose of spiral CT was calculated to be approximately three times higher than that of IVU [12]. However, it should be noted that the radiation dose of IVU will be higher if extra films are taken. Other limitations of spiral CT are its high cost, and high workload, which makes it difficult to accept such cases for emergency examination. Because of these limitations, and the relatively high specificity and positive predictive value of US and IVU, we believe that it would be more realistic to use spiral CT when US and IVU fail to show a reason for the renal colic. References
1. Smith RC, Rosenfield AT, Choe KA, Essenmacher KR, Verga M, Glickman MG, Lange RC (1995) Acute flank pain: comparison of non-contrast-enhanced CT and intravenous urography. Radiology 194: 789794 2. Sommer FG, Jeffrey RB Jr, Rubin GD, Napel S, Rimmer SA, Benford J, Harter PM (1995) Detection of ureteral calculi in patients with suspected renal colic: value of reformatted noncontrast helical CT. AJR 165: 509513 3. Federle PM, McAninch JW, Kaiser JA, Goodman PC, Roberts J, Mall JC (1981) Computed tomography of urinary calculi. AJR 136: 255258 4. Pollack HM, Arger PH, Banner MP, Mulhern CB, Coleman BG (1981) Computed tomography of renal pelvic filling defects. Radiology 138: 645651 5. Smith RC, Verga M, McCarthy S, Rosenfield AT (1996) Diagnosis of acute flank pain. AJR 166: 97101 6. Smith CR, Verga M, Dalrymple N, McCarthy S, Rosenfield AT (1996) Acute ureteral obstruction: value of secondary signs on helical unenhanced CT. AJR 167: 11091113 7. Hillman BJ, Drach GW, Tracey P, Gaines JA (1984) Computed tomographic analysis of renal calculi. AJR 142: 549552 8. Newhouse JH, Prien EL, Amis ES Jr, Dretler SP, Pfister RC (1984) Computed tomographic analysis of urinary calculi. AJR 142: 545548 9. Motola JA, Smith AD (1990) Therapeutic options for the management of upper tract calculi. Urol Clin North Am 17: 191206 10. Heneghan JP, Verga M, Dalrymple N, Rosenfield AT, Smith RC (1996) Rim sign in the diagnosis of ureteric calculi at unenhanced CT. Abstracts of the 82nd annual meeting of RSNA Chicago, Illinois, 219 pp 11. Kawashima A, Sandler CM, Boridy IC, Takahashi N, Benson GS, Goldman SM (1997) Unenhanced helical CTof ureterolithiasis: value of the tissue rim sign. AJR 168: 9971000 12. Schulz RJ, Gignac C (1976) Application of tissueair ratios for patient dosage in diagnostic radiology. Radiology 120: 687690