Real-Time Renal Sonography in spinal Cord Injury Patients: Prospective Comparison with Excretory Urography

0022-534 7/86/1351-0072$02.00/0

Vol. 135, January Printed in U.S.A.

THE JOURNAL OF UROLOGY

Copyright© 1986 by The Williams & Wilkins Co.

Urological Neurology and Urodynamics REAL-TIME RENAL SONOGRAPHY IN SPINAL CORD INJURY PATIENTS: PROSPECTIVE COMPARISON WITH EXCRETORY UROGRAPHY KISHORE G. RAO,* ROBERT H. HACKLER, RAY M. WOODLIEF, MARK N. OZER AND WILLIAM R. FIELDS From the Departments of Radiology, Urowgy and Spinal Cord Injury, McGuire Veterans Administration Medical Center, Richmond, Virginia

ABSTRACT

A prospective blinded comparison of real-time renal sonography and excretory urography was done in 202 urologically asymptomatic patients with spinal cord injury who underwent periodic evaluation. Sonography identified 31 of 31 renal masses (100 per cent) (30 cysts and 1 xanthogranulomatous pyelonephritis), whereas excretory urography identified 14 of 31 masses (45 per cent). Of the 398 kidneys evaluated hydronephrosis owing to an obstructive etiology was noted in 7, all of which (100 per cent) were identified on excretory urography and 6 (86 per cent) were identified on sonography. Only 12 of the 48 kidneys (25 per cent) with typical changes of chronic pyelonephritis on excretory urography were diagnosed correctly by ultrasound. Sonography identified 18 of 23 kidneys (78 per cent) with calculi compared to 20 of 23 (87 per cent) by excretory urography. Although 127 abnormalities were noted in 202 patients, only 21 dictated a change in management. Thirteen abnormalities were visible on a plain film of the kidneys, ureters and bladder (3 kidneys with stones, 1 ureteral stone and 9 bladders with stones). We conclude that sonography and excretory urography are excellent diagnostic modalities for the evaluation of the kidneys. Sonography, plain radiograph of the abdomen and post-contrast injection x-rays on excretory urography frequently offer complementary diagnostic information. The noninvasive nature of ultrasound examination and lack of x-ray exposure combined with no need for special patient preparation make ultrasound examination extremely attractive in this patient population. It is recommended that real-time renal ultrasound and a plain radiograph of the abdomen be used on an alternate basis with excretory urography for the routine followup of spinal cord injury patients. Periodic urological surveillance is necessary in the long-term management of spinal cord injury patients because 1) renal failure remains a leading cause of death, 2) the interplay of factors producing renal deterioration (pyelonephritis, reflux, hydronephrosis, stones and amyloidosis) often are asymptomatic and 3) the ever changing status of the neurogenic bladder eventually may affect the kidneys adversely. 1 Most spinal cord injury centers recommend annual evaluation of the urinary tract, even though no hard data exist as to the optimal time frame. The excretory urogram (IVP), which has been the standard urological test for many years, carries a certain degree of risk to the patient. Conflicting evidence exists that spinal cord injury patients possess more allergic reactions to IVP contrast material compared to the general patient population. 2• 3 Many spinal cord injury patients probably indirectly refuse an IVP by stating to be allergic because of the unpleasant side effects. Because renal sonography was being used frequently in our hospital in spinal cord injury patients, we undertook this prospective evaluation to determine the role of real-time renal

ultrasound in asymptomatic spinal cord injury patients and to compare the diagnostic yield of renal sonography to that of an IVP. MATERIAL AND METHODS

At the time of routine annual IVP, renal sonography was done prospectively on 203 consecutive patients. All of the patients were urologically asymptomatic at the time of examination. Of the 203 patients 6 had undergone previous unilateral nephrectomy. Sonography could not be performed in 1 large patient owing to inability to penetrate through the soft tissues. A total of 202 patients with 398 kidneys was used for the purpose of analysis. There were 200 male and 2 female patients between 20 and 75 years old. Of the patients 106 had lower extremity paralysis or paresis and 96 had paralysis or paresis of all 4 extremities. The duration of injury ranged from a few months to 40 years. Verbal consent was obtained from the patients before sonography. Sonography was performed just before the IVP on the same day in all but 8 patients. In the remaining patients the 2 studies were done within a week of each other. The kidneys were evaluated with the patient supine, or turned to the right or left side using commercially available real-time sector scanners. Additional views with a B arm scanner were taken in

Accepted for publication July 17, 1985. Read at annual meeting of American Urological Association, Atlanta, Georgia, May 12-16, 1985. * Requests for reprints: Department of Radiology, McGuire Veterans Administration Medical Center, 1201 Broad Rock Rd., Richmond, Virginia 23249. 72

73 some patients. Transverse, coronal of each were obtained with a 3 in thin patients in whom a 5 MHz. transducer was used. Sonographic evaluation was confined to the kidneys only. No attempt was made to evaluate the ureters or bladder routinely. The examinations were performed by a technologist or a resident physician under the direct supervision of a staff radiolothe staff gist. The examinations then were interpreted radiologist without the knowledge of the IVP, or other current or previous x-ray studies. The hard copies of the sonogram then were put in a temporary film jacket so that they were not available for the physician interpreting the IVP until after the performance and interpretation of the study. Patient preparation before the IVP consisted of a bowel preparation and overnight fasting. An IVP was done using 100 cc diatrizoate meglumine and diatrizoate sodium. Contrast material was injected intravenously after the initial plain film of the kidneys, ureters and bladder, and a tomogram of the renal area. After the contrast material injection, at least 3 tomographic cuts of the kidney area were taken, followed by a 5-minute cone down view of the kidneys, 10 to 15-minute oblique views of the kidneys and ureters, and a 15 to 20-minute full supine film of the kidneys, ureters and bladder. Patients with Foley catheters did not have the catheter clamped during the examination so as not to cause temporary ureteral dilatation and prevent unpleasant side effects secondary to bladder distension.4 The IVPs were interpreted by a staff radiologist. Previous x-ray studies, if any, were available for comparison at the time of interpretation of the IVP. All of the sonograms, IVPs and patient charts were reviewed. Followup sonography and/or an IVP was done on the majority of the patients approximately 1 year after the initial examination. The results were correlated with other studies, if any, such as plain renal tomograms, cystourethrography, radionuclide study and computerized tomography (CT) to arrive at a final diagnosis. RESULTS

The over-all quality of the examination was judged to be good or satisfactory in 393 of the 398 kidneys (98.7 per cent) on a sonogram as well as on the IVP. In 29 kidneys (15 patients) specific comments were made on the IVP report regarding the presence of excessive bowel gas and feces, motion from extremity spasm or breathing. However, these were placed under the satisfactory category for over-all quality of the examination, most likely on the basis of adequate renal tomograms. Evaluation of 5 kidneys was suboptimal on sonography and IVP to marked renal scarring, stones, poor renal function, overlying bowel gas and/or fecal material. Altogether, 127 potentially significant findings were noted in the 202 patients 1). Many demonstrated more than 1 finding. Among these abnormalities 21 dictated a change in the immediate management of the patients, including renal mass in 1, renal obstruction in 6, kidneys with calculi in 3, bladders with calculi in 9, distal ureteral stone in 1 and gallbladder with stones in 1. Of the 21 abnormalities 13 (3 kidneys with stones, 1 ureteral stone and 9 bladders with stones) were identified on the plain radiographs on the basis of the calcification. TABLE

19,ertal niass. 1'here ~l1e:re 35 r,enal n1-asses u"'x,;L,v"'"v. in 28 Of these masses 30 were cysts on and further confirmatory tests were not done. There were 17 cysts diagnosed on sonography only, 12 of which were 2 cm. or smaller in size. The remaining 5 cysts were up to 4 cm. in size. Of the 5 indeterminate or solid masses on sonography 1 was proved to be xanthogranulomatous pyelonephritis at operation (fig. 1) and 3 were determined to be false positive on the basis of a negative IVP and a radioisotope scan. An additional patient with an indeterminate mass on the sonogram and a normal IVP did not undergo radionuclide study of the kidney because IVPs 1 and 2 years earlier were normal. The IVP demonstrated 17 renal masses in 14 kidneys; 11 were diagnosed as cysts and confirmed on sonography as well. Of the 6 indeterminate or solid masses on an IVP 2 were cysts on sonography and 1 was xanthogranulomatous pyelonephritis. There were 3 false positive scans as determined by normal sonography in all 3 cases, negative radioisotope scan and followup IVP in 2, and a negative CT scan in 1. Chronic pyelonephritis. Chronic pyelonephritis was suspected or diagnosed on sonognphy in 16 kidneys. In 10 cases the findings correlated with those reported on the IVP. Of the remaining 6 kidneys 4 demonstrated lobulation of contour of the kidneys on an IVP without caliceal blunting. In 2 kidneys the IVP revealed typical changes of chronic pyelonephritis in retrospect. There were 50 kidneys in which a diagnosis of chronic pyelonephritis was made on the IVP, including 46 that demonstrated classical changes of focal caliceal blunting associated with parenchymal thinning. In the remaining 4 kidneys other diagnostic possibilities, such as caliceal diverticulum and old vascular insult, were consideted. Of the 48 kidneys with chronic pyelonephritis 46 (96 per cent) were diagnosed correctly on an IVP, whereas 12 were identified on sonography (25 per cent). Hydronephrosis. The term hydronephrosis was used to indicate fullness or dilatation of the calices and the renal pelvis. The diagnosis of hydronephrosis was made prospectively in 33 kidneys on sonography and in 21 on an IVP (table 3). The kidneys were graded subjectively as having mild (grade I) or moderate (grade II) hydronephrosis. No kidney had severe hydronephrosis. Both studies revealed hydronephrosis in 11 kidneys. All of the kidneys with moderate hydronephrosis on an IVP also were detected by sonography. There were 7 kidneys with the final diagnosis of obstruction; 2 cases were caused by radiopaque stones that were removed surgically. A slight of ureteropelvic junction narrowing causing mild hydronephrosis was noted in 1 kidney. In 4 kidneys the hydronephrosis was secondary to obstruction at the bladder level. These returned to normal following false negative sonogram oc-.,.---· ·,.··-,a·- who was on condom cathu,a:,rn,a,;;;;. ;:,onog-raiin:'I demonstrated normal appearance of TABLE

2. Accuracy of sonogram and !VP in diagnosis of renal masses

(398 kidneys) Prospective Diagnosis

Total No. Kidneys

l. Diagnostic yield from screening sonogram and /VP No.

Renal cysts Solid renal mass Renal obstruction Kidneys with chronic pyelonephritis Kidneys with calculi Bladders with calculi Distal ureteral stone Aortic aneurysm Gallbladders with stones

30 1 7 48 23 9

1 1 7

Diagnosed on Sonogram

Diagnosed on IVP

30 1

13 1 7 46

6

12 18 Not evaluated Not evaluated Not evaluated 6

20 9 1 1 1

Final Diagnosis No Mass

Cyst

Solid Mass

23 (30 cysts) 0

0 0 1

13§ (17 cysts) 8 (11 cysts) 2

0 0 1

Sonogram* None Cyst Solid

370 23 5

370 0

None Cyst Solid

384 8 6

371

4t

0

IVP+ 0

3t

* Sensitivity 24/24 = 100 per cent, specificity 370/374 = 99 per cent.

t False positive.

:j: Sensitivity 11/24

§ False negative.

= 46 per cent, specificity 371/374 = 99 per cent.

74

RAO AND ASSOCIATES

FIG. 1. Xanthogranulomatous pyelonephritis. A, longitudinal sonogram of left kidney demonstrates large left upper pole mass (arrows) with solid and cystic components. B, renal tomogram obtained during IVP demonstrates left upper pole mass (arrows) causing caliceal displacement. TABLE

3. Accuracy of sonogram and IVP in diagnosis of renal obstruction (398 kidneys) Final Diagnosis

Prospective Diagnosis of Hydronephrosis

Total No. Kidneys

None Mild Moderate

365 29 4

None Mild Moderate

377 15 6

Obstruction

No Obstruction

Probable No Obstruction

Sonogram* 1t 5 1

364 21:j: 2:j:

0

3:j: 1:j:

!VP§

* Sensitivity 6/7

t False negative. :j: False positive. § Sensitivity 7/7

0 3 4

377 2:j: 0

0

10:j: 2:j:

= 86 per cent, specificity 364/391 = 93 per cent. = 100 per cent, specificity 377/391 = 96 per cent.

the kidney on the right side and mild hydronephrosis on the left side. On the other hand, an IVP demonstrated mild hydronephrosis on the right side and moderate hydronephrosis on the left side. Following the IVP the patient was put on continuous bladder drainage and a repeat IVP a few days later showed complete resolution of bilateral hydronephrosis. False positive sonography findings occurred in 23 kidneys: 21 demonstrated grade I and 2 demonstrated grade II hydronephrosis. An IVP in all except 1 of these patients showed no evidence of hydronephrosis. Followup sonograms or IVPs were done on 21 kidneys and revealed no evidence of hydronephrosis. There were 2 false positive IVPs for obstruction: 1 in a patient with bilateral vesicoureteral reflux, which was demonstrated on a cystourethrogram done shortly after the IVP, and 1 in a patient with the history of recurrent urinary tract infection. Followup IVPs in both patients were normal. An additional 12 kidneys demonstrated hydronephrosis on an IVP; 4 were positive on sonography as well. In all of these kidneys the appearance remained unchanged for at least 1 year (2 years in 4 kidneys and 3 to 8 years in 7). Chronic pyelonephritis contributed to the pelviocaliceal fullness in at least 4 of these kidneys. All of the patients in this group had been paralyzed for at least 10 years and the majority had undergone a previous operation for renal calculi or urinary diversion. There also was no significant alteration of renal function. Therefore, these kidneys were presumed to be nonobstructed. Two patients (2 kidneys) in this group had obstruction owing to calculi during the followup. In both instances sonography or an IVP revealed hydronephrosis and definite change compared to a previous examination. Isolated distal ureteral dilatation without hydronephrosis observed on an IVP occurred in 7 instances (4 patients). Cys-

tourethrography in these patients failed to reveal vesicoureteral reflux. In all of these patients the appearance remained unchanged during at least 2 years. Renal calculi. Of the 398 kidneys l was excluded because stones seen on sonography failed to visualize on an IVP and definite determination could not be made owing to inadequate followup. Stones were diagnosed in 20 kidneys during sonography (table 4). The diagnosis was correct in 18 kidneys, all but l of which had radiopaque calculi. Among these kidneys 15 were diagnosed correctly by both studies. There were 2 false positive interpretations of calculi on sonography as determined by a normal plain radiograph of the abdomen, renal tomograms without contrast material injection and absence of lucent filling defects following the contrast material injection on an IVP. In 5 kidneys the sonogram was falsely negative for calculi (all radiopaque calculi). Calculi were diagnosed on an IVP in 22 kidneys and confirmed in 20. A false positive IVP occurred in 1 patient (2 kidneys) with bilateral hydronephrosis who demonstrated lucent filling defects in the calices bilaterally, which were interpreted as possibly representing lucent calculi. Sonography on the same day failed to demonstrate calculi in the kidneys. The patient was put on continuous bladder drainage via a Foley catheter, and a repeat IVP a few days later showed subsidence of hydronephrosis with no evidence of abnormal filling defects in the calices or renal pelves. In retrospect the apparent filling defects may have been caused by poor mixing of opacified urine or urinary sludge. An IVP was falsely negative for calculi in 3 kidneys: in 2 the calculi were opaque but obscured by the overlying bowel gas and in 1 the calculi were radiolucent. Additional findings. Additional significant findings were seen on the sonogram in 6 patients who demonstrated gallstones and on an IVP in 12 patients (9 had bladder calculi, 1 had a distal ureteral stone, 1 had gallstones and 1 had an abdominal

TABLE 4.

Accuracy of sonogram and IVP in diagnosis of renal calculi (397 kidneys)

Prospective Diagnosis

Total No. Kidneys

Final Diagnosis Calculi

No Calculi

Sonogram* Present Absent

20 377

Present Absent

22 375

2t

18 5:j:

372

20 3:j:

372

!VP§

2t

* Sensitivity 18/23 = 78.2 per cent, specificity 372/374 = 99.5 per cent. t False positive. :j: False negative. § Sensitivity 20/23

= 86.9 per cent, specificity 372/374 = 99.5 per cent.

REl~AL SON-OGRAPHY 11'1 SI="II~J!;..L CORD IN.JURY PA'TIENTS

7.5

aortic AH cf these abnormalities on the IVP were identified on the basis of the calcification. DISCUSSION

Even though an IVP is an excellent diagnostic tool the procedure poses several problems. Adequate bowel preparation before the study is difficult to achieve and unpleasant to the patient. The contrast material produces unpleasant symptoms, such as metallic taste in the mouth, nausea, hot flushes and pain secondary to injection in the majority of patients. In addition, a small number of patients have more serious side effects, such as grand mal seizures, bronchospasm, laryngeal edema, shock, cardiac arrest and, rarely, death. 5 A significant number of spinal cord injury patients report these unpleasant symptoms and many refuse a routine IVP examination. Technical advances during the last decade have enabled visualization of the detailed anatomy of the kidneys with ultrasound. The advent of real-time ultrasound has contributed further to the ease, speed and accuracy of evaluation of the kidneys. Subtle differences in the parenchymal texture between the renal cortex and medulla can be imaged routinely. 6 The usefulness of renal ultrasound in the assessment of renal and perinephric abnormalities has been demonstrated. Compared to an IVP, ultrasound certainly is cost-effective, safer (no allergy and no radiation), less traumatic (painless, no dehydration and no bowel preparation) and less time-consuming. However, for sonography to replace completely the IVP as a routine diagnostic study for asymptomatic spinal cord injury patients, the yield should approximate the IVP in patients with lifethreatening problems, such as stones, pyelonephritis and dronephrosis. Owing to the presence of large amounts of gas and fecal material in the bowel in spinal cord injury patients, clear depiction of renal parenchyma often is difficult on an IVP even after tomographic sections are obtained. Therefore, it is not surprising that sonography identified more than twice as many renal masses as did the IVP. Artifacts from gas and feces in the bowel also may result in false positive diagnosis of a solid mass on an IVP, as noted in 2 kidneys in our study. The marked superiority of sonography in this study group also was owing to the fact that almost all of the masses were cysts. Sonography can identify cysts as small as 0.5 cm. in diameter. Although in the majority of the patients renal cysts produce no symptoms and are of no consequence, it is important to make the correct diagnosis so as not to subject the patients to additional unnecessary invasive tests. The IVP is definitely superior to sonography in the diagnosis of chronic pyelonephritis with parenchymal scarring. This is an important diagnostic observation, especially if the patient has not had previous changes on an IVP, because vesicoureteral reflux could be present. Vesicoureteral reflux in this patient population is detrimental to kidney function, and results in chronic pyelonephritis and a high incidence of infected (struvite) stones. Neither an IVP nor sonography is of any significant value in detecting reflux. In 1 series only 5 of 9 patients (56 per cent) with reflux were detected sonography. 7 Therefore, we believe that assessment of vesicoureteral reflux is accomplished best on a voiding cystourethrogram. Currently, we recommend routine voiding cystourethrography and urodynamic studies every 2 years in asymptomatic spinal cord injury patients on external condom drainage. Diagnosis of the presence of hydronephrosis probably is the single most important determination to be made during the urological evaluation of spinal cord injury patients. Ultrasound is a sensitive method to identify hydronephrosis. 8 ' 9 The relationship between obstruction and dilatation of the urinary system is complex, similar to that in the biliary system. Marked obstruction to the flow of urine may be present in a kidney demonstrating no or mild hydronephrosis. 10 On the other hand, significant dilatation may be present without any obstruction. Whereas sonography demonstrates the renal anatomy, an IVP

FIG. 2. Renal obstruction secondary to calculus. A, sonogram demonstrates evidence of hydronephrosis (arrows) on coronal section of left kidney. B, tomography after contrast material injection during IVP demonstrates nonvisualization of calices of left kidney. Note opaque calculus (arrow) at level of renal pelvis. C, delayed film several hours later demonstrates hydronephrosis on left side.

demonstrates the same under physiological stress because of the strong diuretic property of the contrast material. This difference partly explains the greater sensitivity of an IVP over sonography in the diagnosis of obstruction. In the face of an early or mild obstruction the diuresis appears to accentuate the degree of pelviocaliceal dilatation. More importantly, it may demonstrate the dilatation that was not present previously as shown by Laing and associates. 10 Rarely, severe degree of obstruction may cause marked functional impairment resulting in delayed or nonvisualization of the collecting system on an IVP, whereas hydronephrosis may readily be apparent at sonography, as occurred in 1 of our patients (fig. 2). The ability of sonography to image the kidney irrespective of its function, makes it the primary method of investigation of patients with renal failure. Diagnosis of a large number of kidneys with a dilated upper

76

RAO AND ASSOCIATES

collecting system in this group of asymptomatic patients appears to be related partly to the unique patient population. 11 Multiple factors, such as the state of hydration of the patient, diuretic effect of contrast material injection, altered ureteral peristalsis, vesicoureteral reflux, degree of bladder distension and bladder dyssynergia, in addition to more obvious obstructive etiology, such as a calculus, influence the extent of changes observed on sonography or an IVP. Differentiation of transient physiological dilatation from dilatation secondary to obstruction cannot always be made on a single sonogram or IVP. This is true particularly in long-standing spinal cord injury patients with renal scarring and/or postoperative changes. For sonography to be an effective screening study even a slight degree of separation of central pelviocaliceal echo complex should be viewed with suspicion and investigated further. With this objective in mind a sonograhic false positive rate of 7 per cent appears to be acceptable. The presence of multiple renal or staghorn calculi severely limits the visualization of intrarenal anatomy during sonography and may result in a false negative study for hydronephrosis. We agree with Scheible and Talner12 that these patients are best evaluated with an IVP. A few scattered reports exist on the usefulness of ultrasound in the detection of renal calculi. 13• 14 Our study is a prospective comparison of sonography and an IVP in renal calculous disease. Renal calculi in spinal cord injury patients are invariably composed of struvite and caused by urea-splitting bacteria. In many cases the stone growth can be compared to a slow growing malignancy, since stones and/or infection cannot be eradicated completely. Currently, it is imperative to detect stones at an early stage so that therapeutic modalities, such as percutaneous ultrasound lithotripsy and extracorporeal shock wave therapy, can be more effective. Renal stones produce a characteristic appearance of a focal echogenic area with shadowing that is observed readily with real-time equipment (fig. 3). Since observation of these findings is not related to radiopacity of the

stones, radiolucent stones, such as uric acid or matrix, can be differentiated readily from other causes of radiolucent filling defects in the pelviocaliceal system. Although the majority of stones are radiopaque, they are not missed infrequently on plain radiographs owing to the overlying bowel gas and fecal material. We have a number of instances in which the initial plain radiograph of the abdomen failed to demonstrate kidney calculi but the sonogram did. This finding prompted a repeat abdominal radiograph, plain renal tomograms or some additional examination that confirmed the presence of stones. False negative sonography for calculi, which occurred in 5 kidneys in our experience, apparently was owing to 2 reasons. In 2 kidneys the calculi measured 1 to 3 mm. in size and probably were too small to cast an acoustic shadow. In the remaining 3 kidneys they were located in the renal pelvis just medial to the bulk of renal tissue. It is difficult to differentiate the shadowing produced by calculi in this location from normal shadowing produced by the medial portion of the rib or transverse process of the lumbar spine in the absence of a fluid-filled pelvis. In fact, one of our false positive sonograms for calculus presumably was caused by the transverse process of the lumbar spine. It should be emphasized that extrarenal abnormalities detected on sonography were incidental. No effort was made during the sonographic study to evaluate the remainder of the abdomen systematically. Identification of 6 patients with gallstones was due to the fact that the gallbladder often is seen during right renal evaluation. All of the 12 significant extrarenal abnormalities noted on an IVP demonstrated sufficient calcification to be visible on the plain radiograph of the abdomen. From our study it is apparent that the yield of findings that influence the immediate management of routine spinal cord injury patients is low. Since more than 50 per cent of these findings were visible on the basis of their calcification on the plain radiograph of the abdomen, continued use -0f an abdominal radiograph appears to be mandatory. Sonography and the IVP identify different renal abnormalities with varying degrees of accuracy. In this patient population sonography is superior in the diagnosis of renal masses and inferior in the detection of chronic pyelonephritis. Sonography and the IVP have comparable diagnostic yields in the diagnosis of renal calculi. Even though sonography is sensitive the IVP remains the gold standard in the detection of renal obstruction. Combination of the data from both studies vastly diminishes the incidence of false positive and false negative results. Therefore, we recommend ultrasound and an IVP for the initial routine evaluation of the kidneys in spinal cord injury patients. Ultrasound and a plain film of the kidneys, ureters and bladders then may be used on an alternate basis with an IVP for followup evaluation. Abnormalities noted on 1 study may be clarified with the other or with additional studies as appropriate. REFERENCES 1. Hackler, R. H.: A 25-year prospective mortality study in the spinal

2. 3. 4. 5.

FIG. 3. Renal calculus. A, parasagittal section of sonogram through right kidney demonstrates echogenic focus with shadowing in lower pole (arrows). B, plain radiograph demonstrates calculus in lower pole of right kidney (arrow).

6. 7.

cord injured patient: comparison with the long-term living paraplegic. J. Urol., 117: 486, 1977. Simpson, K. H., Cottrell, T. L. C. and Lloyd, F. A.: Intravenous pyelography in the spinal cord injury patient. Amer. J. Surg., 103: 217, 1962. Scher, A. T.: Are spinal cord injury patients more prone to contrast reactions? Paraplegia, 16: 366, 1979. Barbaric, Z. L.: Autonomic dysreflexia in patients with spinal cord lesions: complication of voiding cystourethrography and Heal loopography. Amer. J. Roentgen., 127: 293, 1976. Witten, D. M., Hirsch, F. D. and Hartman, G. W.: Acute reactions to urographic contrast medium: incidence, clinical characteristics and relationship to history of hypersensitivity states. Amer. J. Roentgen., 119: 832, 1973. Rosenfield, A. T., Taylor, K. J. W., Crade, M. and DeGraaf, C. S.: Anatomy and pathology of the kidney by gray scale ultrasound. Radiology, 128: 737, 1978. Calenoff, L., Neiman, H. L., Kaplan, P. E., Nanninga, J.B., Brandt,

[NJTIRY PA'I1IEf-·JTS

80

90

100

lL 120

130

140

T. Do and Hamilton, R R: Urosonography in spinal cord injmy patientso Jo UroL, 128: 1234, 19820 Ellenbogen, Po Ho, Scheible, Fo WO, Talner, Lo R and Leopold, Go R: Sensitivity of scale ultrasound in detecting urinary tract obstruction" Amero 130: 731, 197R lVl'JHCCalll, Go, Cozzupoli, po and Maggiore, Qo: The value of as a screening prncedure for urological disorders in renal failure" Jo 130: 19830 Laing, Fo Co, Jeffrey, RR, JL and Wing, V WO: versus excretory urography in evaluating acute flank pain" Radiology, 154: 613, 19850 Irvine, Ao R: Upper urinary tract dilatation in paraplegia" Brit Jo UroL, 31: 47, 19590 Scheible, Ao and Talner, Lo R: Gray scale ultrasound and the genitourinary tracto A review of clinical applications" Rado Clino No AmeL, 17: 281, 19790 Pollack, ff Mo, Arger, Po R, Goldberg, R R and Mulholland, So Go: Ultrasonic detection of nonopaque renal calculL Radiology, 127: 233, 19780 Edell, So and Zegel, R: Ultrasonic evaluation ofrenal calculi" Amero Jo Roentgeno, 130: 261, 19780 EDITORIAL COMMENT

The concept that ultrasonography may be as informative as an IVP in patients with spinal cord injury is not new, and has been noted by Brandt and associates in 1981, 1 and by Calenoff and associates in 1982

77

(reference 7 in article)" The authors have expanded considerably these earlier observations, and are to be commended for the design, magnitude and thoroughness of their study" They suggest that an IVP and may be used interchangeably, each on a biannual basis and at same time call attention to some of the pitfalls inherent in the use of renal ultrasonographyo To these pitfalls might be added what many regard as the greatest shortcoming of all, namely the extreme operator dependency of ultrasound. It is unfortunate but true that the quality of ultrasonograms can vary greatly from l institution to anotheL In view of this fact, recommendations that ultrasonography replace other modalities must be coupled with the admonition that such recommendations imply sonography of the highest quality" Anything less than this will lead to unacceptably poor-and possibly direresultso The authors wisely indicate that the IVP, although much maligned lately, 2 still is the gold standard to detect obstruction as well as caliceal scarring" As far as I can see, their study does not argue for a further shift away from this examination" Rather, it sensibly suggests alternating it with a nonx-ray technique in patients requiring longterm, regular radiological followup.

Howard Mo Pollack 531 Ashmead Road Cheltenham, Pennsylvania L Brandt, To Do, Neiman, R Lo, Calenoff, L., Greenberg, Mo, Kaplan, Po K and Nanninga, Jo R: Ultrasound evaluation of the urinary system in spinal-cord-injury patients" Radiology, 141: 473, 198L 20 Pollack, Ho Mo and Banner, Mo P.: Current status of excretory urographyo Urol. Clino K Amero, in press"