Limitations of Ultrasonography in the Evaluation of Urolithiasis: A Correlation With Computed Tomography

Abstract

Background and Purpose:

Cumulative radiation exposure from imaging studies is hazardous. In chronic diseases such as urolithiasis, efforts are made to limit radiation exposure, particularly for routine surveillance. We sought to determine the correlation of ultrasonography (US) compared with noncontrast CT (NCCT) in detecting and determining size of stones.

Patients and Methods:

Findings were evaluated in patients who underwent both imaging modalities within a 90-day period between July 2008 and June 2010. Urinary calculi were noted on NCCT in 72 patients. The sensitivity of US to determine the number, size, and location of the stones as described on official radiology reports were compared in reference to NCCT.

Results:

There were 203 urinary calculi in 90 urinary tracts identified on NCCT imaging. The sensitivity, specificity, and accuracy of detecting specific stones on US were 40%, 84%, and 53%. Correlation between US and NCCT findings decreased with smaller stone size and ureteral location and increased with right-sided laterality. For identified stones, larger stone size discrepancies were noted in up to one-third of stones on US.

Conclusions:

Despite concern for excessive radiation exposure, urologists should recognize limitations of US in the evaluation of urolithiasis. As the ideal study to image stones, particularly for routine surveillance, remains unclear, tese data also supports the need for low-dose NCCT protocols and/or selective use of alternative modalities, such as magnetic resonance urography.

Introduction

U rolithiasis is one of the most prevalent and recurrent diseases treated by urologists. The diagnosis is based on imaging of the urinary tract to determine the size, location, number, and significance of the stones. Noncontrast CT (NCCT) is considered the gold-standard imaging modality because of its high sensitivity and specificity in the identification of both renal and ureteral calculi.^1

–5 Given the incidence of first-time stones in the evaluation of acute flank pain as well as recurrent stone episodes, however, many affected patients will undergo multiple studies over their lifetime.

Recent reports have reiterated the risks of cumulative radiation exposure from imaging studies such as NCCT.^6,7 In response, increased efforts are made to use nonionizing modalities, such as ultrasonography (US), particularly for routine surveillance. We sought to determine the accuracy of US compared with NCCT in detecting and determining the characteristics of urinary calculi in a cohort of stone patients.

Patients and Methods

After Institutional Review Board approval, we retrospectively analyzed patients who underwent NCCT and US imaging within a 90-day period between June 2009 and September 2010. Indications for imaging included acute flank pain, hematuria, or history of nephrolithiasis. Imaging in patients with a history of stone passage or treatment during the 90-day period was excluded. Helical NCCT of the abdomen and pelvis was formatted into 3- to 5-mm contiguous axial images. Dedicated renal or abdominal US was performed using grey scale sonography with attention to the urinary tract in multiple anatomic planes. Parameters reviewed included: Date of studies; number, size, and location of stones noted on official radiologic interpretation; secondary radiographic findings on US including hydronephrosis or absence of a ureteral jet; demographics; and clinical outcomes.

Sensitivity analyses for the detection of stones on US were calculated using NCCT as the standard reference. Stones were grouped based on longest axis diameter using clinically relevant cutoffs (≤4 mm, 4.1–9.9 mm, ≥10 mm) and location (renal, ureteral).^8
–10 Stones described as punctate foci on imaging were considered as ≤4 mm. Cases in which the number or size of stones were otherwise indeterminate on the official report were excluded.

The relationships between imaging correlation and body mass index (BMI), laterality, and time between studies were determined using discriminant and chi-square analyses for continuous and categorical variables. A P value of ≤0.05 was considered statistically significant.

Results

Of 130 patients (142 urinary tracts) who underwent NCCT and US, 72 (90 urinary tracts) were noted with 203 urinary calculi on NCCT. Patients with stones on NCCT included 44 males and 28 females with a mean age of 47 (range 19–77). Mean BMI was 30 (range 20–51). Right urinary tracts were involved in 51 (57%). The mean interval between NCCT and US was 34±30 days. On NCCT, multiple stones (mean 2.3±1.6) were noted in 54 (60%) urinary tracts. Mean stone size was 4 mm±1.6 and located in the kidney (74%), ureter (9%), or both (17%).

In patients with calculi on NCCT, at least one stone was identified in 52 urinary tracts on US (Table 1). The sensitivity, specificity, and accuracy of US for detecting any stone were 58%, 91%, and 73%. Specific visualization of corresponding stones was noted in 81 of 203 stones. The sensitivity, specificity, and accuracy of US for detecting an individual stone were 40%, 84% and 53%. Identification of stones on US decreased with smaller stone size and ureteral location (Table 2). The sensitivity for calculi ≤4 mm was 35%, 4.1–9.9 mm was 48%, and ≥10 mm was 86%. Overall, 26 (13%) of 122 stones absent on US were >4 mm. The sensitivity for detecting ureteral stones (15%) was lower than for renal stones (44%). Evaluation of the ureter improved to 67%, however, with consideration of secondary characteristics on US.

Table 1.

Detection of 203 Urinary Calculi in 90 Urinary Tracts on Noncontrast Computed Tomography

Detection of any stone	NCCT		Sensitivity of US	Specificity of US	Accuracy of US	Positive predictive value of US	Negative predictive value of US
US	Positive	Negative
Positive	52	7	58%	91%	73%	88%	64%
Negative	38	68
Detection of specific stone
Positive	81	13	40%	84%	53%	86%	36%
Negative	122	68

NCCT=noncontrast computed tomography; US=ultrasonography.

Table 2.

Detection of 203 Urinary Stones on Ultrasonography According to Size and Location

Stone size (mm)	Total # of stones	# Detected	# Missed	Sensitivity (%)
≤4.0	148	52	96	35
4.1–9.9	48	23	25	48
≥10	7	6	1	86
Stone location
Renal	176	77	99	44
Renal +2 degree features	176	77	99	44
Ureter	27	4	23	15
Ureter +2 degree features	27	18	9	67

Discordance in stone group size was observed in 29 (36%) of 81 identified calculi (Table 3). Measurement of stones on US compared with NCCT resulted in a higher group classification in 25 (31%) cases. In addition, 13 (14%) of 94 echogenic foci identified on US as calculi were not observed on NCCT. Seven (54%) of these foci measured >4 mm. Of four of these patients with follow-up NCCT, no corresponding calculi were observed. Thirty-four (47%) of the 72 patients noted with stones on NCCT underwent endoscopic treatment (30 unilateral/3 bilateral ureteroscopies, 1 percutaneous nephrolithotomy) based on their imaging. Operative findings were consistent with NCCT imaging in all cases and in 16 of 37 (43%) cases on US. In the remaining 21 cases, 12 patients were found with renal stones not observed on US and 2 patients were noted with smaller sized stones than described on US. Of patients with ureteral stones on operative findings, none were noted on US and only 8 (57%) had secondary characteristics on US suggestive of pathology.

Table 3.

Agreement in Stone Size Between Noncontrast Computed Tomography and Ultrasonography in 81 Cases

	Stone size on US (mm)
Stone size on NCCT (mm)	≤4.0	4.1–9.9	≥10
≤4.0	35	16	1
4.1–9.9	2	13	8
≥10	0	2	4

US=ultrasonography; NCCT=noncontrast computed tomography.

No difference in detection of calculi between studies was observed in terms of BMI. Right-sided laterality had a positive correlation with detection of renal stones even ≤4 mm. Time between imaging did not adversely affect the association between US and NCCT (Table 4).

Table 4.

Influence of Patient Factors on Agreement in Stone Detection Between Noncontrast Computed Tomography and Ultrasonography

	Right side			BMI		Days between imaging studies
	Chi-square value	Correlation	P value	Correlation	P value	Correlation	P value
Any stone	0.422	0.066	0.516	0.124	0.314	0.018	0.863
Specific stone	5.12	0.154	0.024	0.052	0.511	0.13	0.057
Stone ≤4 mm	6.174	0.204	0.013	0.103	0.291	0.144	0.082
Stone 4.1–9.9 mm	0.071	−0.038	0.79	0.03	0.856	0.33	0.022
Stone ≥10 mm	1.556	0.471	0.212	0.402	0.507	0.325	0.478
Renal stone	4.382	0.158	0.036	0.029	0.744	0.131	0.083
Ureteral stone	0.006	−0.015	0.936	0.196	0.451	0.282	0.155
Ureteral stone +2-degree features	0.074	−0.052	0.785	0.054	0.837	0.188	0.349

Analysis: Chi-square test (right side) and canonical correlation (BMI and days between imaging studies).

BMI=body mass index.

Discussion

Urolithiasis develops in up to 12% of the population in their lifetime.¹¹ A patient with a single stone episode is thought to have a 10% to 23% rate of recurrence per year that cumulates to 50% at 5 years.^12,13 This results in patients undergoing multiple repeated studies for the evaluation of acute symptoms as well as surveillance of stone disease. In our study, we confirmed that US has a limited role in the diagnosis and characterization of calculi throughout the urinary tract compared with NCCT. The sensitivity, specificity, and accuracy of detecting specific stones on US were 40%, 84%, and 53%. Smaller stone size and ureteral location were associated with discordance between US and NCCT findings while right-sided laterality improved correlations. For stones identified on US, increased stone size discrepancies were noted in up to one-third of cases.

NCCT is the current gold standard for the diagnosis of urinary calculi with reported sensitivities and specificities of 94% to 97% and 96% to100%.^1

–5 In comparison, the sensitivities for stone detection on plain abdominal radiography or intravenous urography are 45% to 58% and 52% to 87%.^2,3,14,15 With less influence by body habitus and degree of bowel gas, NCCT can reliably obtain anatomic detail of all stones types, except for matrix or indinavir-like compositions, and discriminate between differences in attenuation between the renal pelvis and parenchyma.^16,17 Based on phantom studies, the margin of error on NCCT with actual stone size is <3.6%.¹⁸ In addition, information regarding stone density to predict composition and treatment response as well as diagnosis of extraurinary pathology can be obtained. Functional assessment, however, needs the addition of intravenous contrast and volume reconstruction with delayed imaging.¹⁹

The primary concern with repeated routine NCCT imaging in stone disease remains the associated cumulative radiation exposure in a relatively young patient population. The International Commission on Radiation Protocol recommends that exposure not exceed 20 mSv per year during a 5-year period or 50 mSv in any single year.²⁰ The median effective radiation dose with a single NCCT of the abdomen and pelvis is 15 mSv (interquartile range 10–20). Using risk models, it has been estimated that approximately 1 in 1400 patients age 60 undergoing NCCT would develop a radiation-induced solid cancer or leukemia. This ratio increases to 1 in 1000 patients age 40 and is of particular concern given the age of peak stone incidence is 40 to 49 years.^7,21 Katz and associates²² noted that a subset of patients with a history of nephrolithiasis underwent three or more examinations at their hospital over a 6-year period with estimated effective doses of 19.5 to 153.7 mSv. Similarly, Ferrandino and colleagues²³ reported a mean of four (1.7 NCCT) radiographic studies per patient in 1 year after an acute stone episode. In this multi-institutional study, the median total effective radiation dose was 29.7 mSv per patient with 20% of the cohort receiving >50 mSv.

US is an attractive urinary tract imaging modality that needs no radiation exposure or risk of subsequent attributable malignancy. It is a quick, inexpensive, repeatable study that is particularly useful in pregnancy and for the diagnosis of asymptomatic obstruction after ureteroscopy.^19,24,25 Similar to other groups, however, we noted a decreased utility with US compared with NCCT in the evaluation of urolithiasis.^3
–5 In our study, the sensitivity of US in the detection of any stone or a specific stone was only 58% and 40%. In an evaluation of 101 calculi in 39 patients, Fowler and coworkers⁴ reported sensitivity values for US in similar settings of 44% and 24%. The low sensitivity of US compared with NCCT suggests that it has poor utility for the exclusion of urolithiasis.

Delineation of stones on US often necessitates acoustic shadowing, and a lack of shadowing may occur from intervening tissue with different acoustic impedance values, filling of the shadow from reverberation or incorrect selection of transducer settings.²⁶ Identification of stones ≤4 mm was most limited on US. While the clinical significance of these sized stones has been questioned, the inability to identify small stones prohibits determining the metabolic activity in recurrent stone formers.^8,9

Although patient BMI showed no influence, we did find a positive relationship between right-sided laterality with stone correlation on US. Previous hypotheses have suggested that visualization of the right renal system is easier because of improved acoustic windows from the liver and less need for intercostal scanning because of the inferior position of the kidney.²⁷ As reported in other studies, ureteral stones in our cohort were less recognized on US because of its inability to image the entire ureter.^3,5

The addition of secondary characteristics such as hydronephrosis or absent ureteral jet did improve the identification of obstruction on US. The presence of a normal ureteral jet, however, does not exclude a ureteral stone, and generally additional imaging is needed to confirm and further characterize the cause of obstruction.⁵

We noted a relatively high specificity for US compared with NCCT (84%–91%) in the identification of stones reflecting that calculi found on US are likely present on NCCT. Of stones seen on US, however, 31% were upsized in maximal diameter and noted in two patients operatively. The reproducibility of size on US has been questioned because of its reliance on orthogonal planes rather than fixed planes.¹⁹ This can affect treatment decisions, because stones >4 mm need an increased rate of intervention and stones ≥10 mm may not be candidates for observation.^8
–10 Also, resolution of closely spaced stones and differentiation of parenchymal calculi can be difficult.

In our study, 14% of stones (54% of which were >4 mm) reported on US were not evident on NCCT. Sound refractions at the interface of the renal parenchyma, sinus fat, intrarenal vessels, and collecting system have been implicated as causes for false positive findings that could lead to additional unnecessary surgical interventions.^24,27

There are several limitations with this study. We accepted a maximal interval between studies of 90 days, because this is the shortest period to which we would consider reimaging for stone surveillance. Although we excluded the laterality of patients receiving treatment or after a history of stone passage, concurrent studies on the same day for all patients would have been most optimal. There was no adverse correlation in imaging and time between studies on discriminant analysis, however. We additionally lacked complete clinical follow-up in all patients, which limited definition of true positives and true negatives. In the 47% of patients who underwent surgery, correlation with NCCT findings was 100% but only 43% with US primarily from false-negative findings. Finally, this retrospective study required analysis based on official radiography reporting for stone characterization. Because US accuracy is highest prospectively in real time, however, previous studies have shown improved or no significant difference in US sensitivity between original reporting compared with direct imaging reinterpretation.^4,27

Despite these limitations, the poor utility of US and the high radiation exposure associated with standard NCCT support the need for effective alternative imaging strategies to fully characterize urinary calculi. Several investigators have studied lower electrical current settings (≤30 mAs) for low-dose NCCT protocols compared with standard NCCT settings (140–180 mAs). While the reported sensitivity (95%–96%) and specificity (86%–100%) of low-dose NCCT are promising, the accuracies for small stones and in patients with a BMI ≥30 have been questioned.^28,29

Magnetic resonance urography (MRU) can provide a nonradiation-based comprehensive examination of the soft tissues of the abdomen and pelvis that depicts a signal void within the collecting system and secondary obstructive findings to identify urinary calculi.³⁰ Particularly advocated in pediatrics and pregnancy, MRU has shown comparable stone detection rates with NCCT in limited evaluation, although additional studies are needed.³¹

Conclusions

Despite continued concern for excessive lifetime radiation exposure with NCCT, urologists should recognize limitations of US in the evaluation of urolithiasis. In our cohort, only 40% of stones identified on NCCT were observed on US, and one-third of stones seen on US exhibited significant size differences that could have led to alternative management practices. Correlation between US and NCCT findings decreased with smaller stone size and ureteral location and increased with right-sided laterality. Because the ideal study to image stones particularly for routine surveillance remains unclear, these data also support the need for low-dose NCCT protocols and/or selective utilization of modalities such as MRU.

Footnotes

Acknowledgment

The authors would like to acknowledge Roslin V. Hauck, Ph.D., for her assistance with the statistical analyses.

Disclosure Statement

No competing financial interests exist.

Abbreviations Used

References

Smith

, Verga

, McCarthy

, Rosenfield

. Diagnosis of acute flank pain: Value of unenhanced helical CT. AJR Am J Roentgenol, 1996; 166:97–101.

Miller

, Rineer

, Reichard

et al. Prospective comparison of unenhanced spiral computed tomography and intravenous urogram in the evaluation of acute flank pain. Urology, 1998; 52:982–987.

Yilmaz

, Sindel

, Arslan

et al. Renal colic: Comparison of spiral CT, US, and IVU in the detection of ureteral calculi. Eur Radiol, 1998; 8:212–217.

Fowler

, Locken

, Duchesne

, Williamson

. US for detecting renal calculi with nonenhanced CT as a standard reference. Radiology, 2002; 222:109–113.

Sheafor

, Hertzberg

, Freed

et al. Nonenhanced helical CT and US in the emergency evaluation of patients with renal colic: Prospective comparison. Radiology, 2000; 217:792–797.

Berrington de

González A

, Mahesh

, Kim

et al. Projected cancer risks from computed tomographic scans performed in the United States in 2007. Arch Intern Med, 2009; 169:2071–2077.

Smith-Bindman

, Lipson

, Marcus

et al. Radiation dose associated with common computed tomography examinations and the associated lifetime attributable risk of cancer. Ann Intern Med, 2009; 169:2078–2086.

Miller

, Kane

. Time to stone passage for observed ureteral calculi: A guide for patient education. J Urol, 1999; 162:688–691.

Monga

. Progression of nephrolithiasis: Long-term outcomes with observation of asymptomatic calculi. J Endourol, 2004; 18:534–539.

10.

Preminger

, Tiselius

, Assimos

et al. 2007 guideline for the management of ureteral calculi. J Urol, 2007; 178:2418–2434.

11.

Sierakowski

, Finlayson

, Landes

et al. The frequency of urolithiasis in hospital discharge diagnoses in the United States. Invest Urol, 1978; 15:438–441.

12.

Churchill

. Medical treatment to prevent recurrent calcium urolithiasis. A guide to critical appraisal. Miner Electrolyte Metab, 1987; 13:294–304.

13.

Ljunghall

. Incidence of upper urinary tract stones. Miner Electrolyte Metab, 1987; 13:220–227.

14.

Levine

, Neitlich

, Verga

et al. Ureteral calculi in patients with flank pain: Correlation of plain radiography with unenhanced helical CT. Radiology, 1997; 204:27–31.

15.

Mutgi

, Williams

, Nettleman

. Renal colic. Utility of the plain abdominal roentgenogram. Arch Intern Med, 1991; 151:1589–1592.

16.

Smith

, Coll

. Helical computed tomography in the diagnosis of ureteric colic. BJU Int, 2000; 86,suppl 1:33–41.

17.

Bruce

, Munch

, Hoven

et al. Urolithiasis associated with the protease inhibitor indinavir. Urology, 1997; 50:513–518.

18.

Olcott

, Sommer

, Napel

. Accuracy of detection and measurement of renal calculi: In-vitro comparison of three-dimensional spiral CT, radiography, and nephrotomography. Radiology, 1997; 204:19–25.

19.

Sandhu

, Anson

, Patel

. Urinary tract stones—Part I: Role of radiological imaging in diagnosis and treatment planning. Clin Radiol, 2003; 58:415–421.

20.

ICRP Publication 60: 1990 Recommendations of the International Commission on Radiological Protection. Annals of the ICRP vol 21/1-3. Elmsford, NY: Pergamon Press, 1991.

21.

Romero

, Akpinar

, Assimos

. Kidney stones: A global picture of prevalence, incidence, and associated risk factors. Rev Urol, 2010; 12:e86–96.

22.

Katz

, Saluja

, Brink

, Forman

. Radiation dose associated with unenhanced CT for suspected renal colic: Impact of repetitive studies. AJR Am J Roentgenol, 2006; 186:1120–1124.

23.

Ferrandino

, Bagrodia

, Pierre

et al. Radiation exposure in the acute and short-term management of urolithiasis at 2 academic centers. J Urol, 2009; 181:668–673.

24.

Rao

. Imaging for kidney stones. World J Urol, 2004; 22:323–327.

25.

Manger

, Mendoza

, Babayan

, Wang

. Use of renal ultrasound to detect hydronephrosis after ureteroscopy. J Endourol, 2009; 23:1399–1402.

26.

King

III , Kimme-Smith

, Winter

. Renal stone shadowing: An investigation of contributing factors. Radiology, 1985; 154:191–196.

27.

Middleton

, Dodds

, Lawson

, Foley

. Renal calculi: Sensitivity for detection with US. Radiology, 1998; 167:239–244.

28.

Poletti

, Platon

, Rutschmann

et al. Low-dose versus standard-dose CT protocol in patients with clinically suspected renal colic. AJR Am J Roentgenol, 2007; 188:927–933.

29.

Jellison

, Smith

, Heldt

et al. Effect of low dose radiation computerized tomography protocols on distal ureteral calculus detection. J Urol, 2009; 182:2762–2767.

30.

Kalb

, Sharma

, Salman

et al.

Acute abdominal pain: Is there a potential role for MRI in the setting of the emergency department in a patient with renal calculi?

J Magn Reson Imaging, 2010; 32:1012–1023.

31.

Regan

, Kuszyk

, Bohlman

, Jackman

. Acute ureteric calculus obstruction: Unenhanced spiral CT versus HASTE MR urography and abdominal radiograph. Br J Radiol, 2005; 78:506–511.