| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters to the Editor |
Schneider Children's Hospital, Long Island Jewish Medical Center, 270-05 76th Avenue, New Hyde Park, NY 11040
Babies and Children's Hospital, New York, NY
Editor:
The article by Dr Lonergan and colleagues in the May 1998 issue of Radiology (1) is an interesting contribution to the diagnostic imaging of acute pyelonephritis. The well-designed and well-executed study showed a sensitivity and a specificity of magnetic resonance (MR) imaging that are equal to and perhaps even greater than those of renal cortical scintigraphy with technetium 99m dimercaptosuccinic acid (DMSA), which is generally considered the standard for determining renal involvement in urinary tract infections. The present study deals with the clinical application of the same authors' previous work on the use of MR imaging in experimental pyelonephritis in piglets, the findings of which were published in Radiology in 1996 (2). In both instances, the best pulse sequence for the demonstration of areas of pyelonephritis at MR imaging proved to be gadolinium-enhanced fast spin-echo inversion recovery. Dr Pennington, one of the co-authors, also reported at the 1997 annual meeting of the Society for Pediatric Radiology that gadolinium-enhanced fast spin-echo T2-weighted images were equally diagnostic as inversion-recovery images.
We had the opportunity to apply MR imaging in the detection of acute pyelonephritis in a small group of children as well, and our findings agree with those of Dr Lonergan and colleagues. Obviously, MR imaging is an emerging, powerful new modality for the evaluation of urinary tract infections, and it may eventually prove of high value when used appropriately and in selected cases.
In a commentary (3) published recently in Pediatric Nephrology, Verrier Jones states that
although [urinary tract infection] and [vesicoureteral reflux] are common conditions, reflux nephropathy . . . is much less common, although potentially more serious. There is no sound evidence that it is necessary or even desirable to subject so many children to intensive radiological investigation. Potential benefits should be carefully balanced against the economic implications and psychological disadvantages of this intensive program.
In the opening paragraph of the article by Dr Lonergan and colleagues (1), the potential consequences of renal scars, which now can be detected in more than half of the children with urinary tract infections (4,5) by using ever-more-sensitive imaging techniques (such as 99mTc DMSA scanning and now MR imaging), are portrayed as a serious epidemiologic threat. If indeed according to these predictions one in five children will develop hypertension and one in 10 will develop end-stage renal disease, one would expect a large number of young adults to be on a waiting list for renal transplantation.
We wish to warn against these dire predictions, with which the recent literature is replete, because there is no sound evidence that childhood pyelonephritis demonstrated at 99mTc DMSA scanning will result in numerous cases of end-stage renal disease.
According to a recent census, the population of the United States in 1995 was approximately 263 million persons, of whom 60 million were women of reproductive age (15–49 years). With 3.9 million births annually, the fertility rate in these women is 65.6 births per 1,000 women. With the low number of expected deaths subtracted, approximately 3.8 million children will reach the age of 15 years each year (6). Assuming a conservative estimate of a 2% prevalence of urinary tract infections, there would be 80,000 urinary tract infections over 15 years in these children. Assuming again an even spread of urinary tract infections over these 15 years, there would be 5,300 cases annually from a single cohort. Even with a purposely chosen low number from the literature of a 50% incidence of scar development, one would expect 2,650 scars annually, or 26,500 scars in 10 years.
According to published statistics, however, there were 45,000 patients in the United States in 1993 who had end-stage renal disease; less than 10% or at most roughly 4,500 of these patients had cases that could be attributed to chronic childhood pyelonephritis (7). Certainly, these cases of end-stage renal disease were gathered over many years, perhaps more than 10. Although the computations are rough estimates and assumptions, they suggest that the danger of end-stage renal disease has been overestimated by a factor of at least 6 by using the minimum of published numbers of risk estimates.
In support of our thesis, Kincaid-Smith (8) stated in an editorial that
the proportion of patients with reflux nephropathy who develop end stage renal failure cannot be easily calculated because in general only patients with symptoms are studied. Follow-up of these detected by screening for bacteriuria (without symptoms) has produced little evidence of progression . . . the highest estimates suggest that no more than five to ten women per million of population present yearly with end stage renal failure from this cause, so that probably only one woman in every 1,000 with reflux nephropathy is at risk of progressing to end stage renal failure.
Data on the incidence of hypertension from childhood pyelonephritis are also unavailable as derived from publications of highly selective cases (9,10).
It is, therefore, more appropriate to concentrate our efforts and resources to identify the specific group of children at risk for progression from urinary tract infection to end-stage renal disease rather than to focus on our increasing ability to detect every minute scar with ever-more-precise imaging. We under no circumstances try to detract from the contributions of renal cortical scintigraphy and MR imaging in the accurate detection of acute pyelonephritis and subsequent renal scar formation; we simply believe that these tests should not be used routinely. Much work is now needed to learn when and for whom they are appropriate.
There are many reasons for our lack of knowledge with regard to the exact relationship of urinary tract infections and renal scars with progressive renal disease. Information on urinary tract infection, vesicoureteral reflux, and reflux nephropathy date from the 1950s and was based mostly on findings of intravenous pyelography (now proved incapable of depicting most scars) obtained with no control subjects and with a limited follow-up. We have no proof that scars shown with intravenous pyelography have any relationship to the small scars now shown on 99mTc DMSA scans (10,11). Also, there is no convincing evidence that therapeutic intervention has any effect on final outcome (10). Thus, specific factors that favor scarring and long-term adverse outcome need to be studied so that imaging may become more selective. Rough estimates indicate that millions and millions of dollars are spent annually on the diagnosis and treatment of urinary tract infections without clear-cut evidence that the cost (and the probable psychologic effect) are justified.
Recent developments indicate that vesicoureteral reflux is genetic. Oddly, its in utero occurrence, as identified by using prenatal and postnatal ultrasonography, defines a certain population at risk: mostly male patients with possible bladder dysfunction. Further research in this area may give a better focus to the imaging and management of urinary tract infections (12,13).
In conclusion, Dr Lonergan and colleagues have shown that MR imaging is a new, accurate imaging modality for the detection of renal involvement and scar formation in childhood urinary tract infections. It should not become a screening test before findings of further studies help delineate particular circumstances for its use. The importance of detection of generally minor changes in the renal parenchyma by using renal cortical scintigraphy or MR imaging or their effect on management have yet to be proved.
References
Department of Radiologic Pathology, Armed Forces Institute of Pathology, Building 54, Room M 121, 14th and Alaska Northwest, Washington, DC 20306-6000
Department of Radiology, Riley Hospital for Children, Indianapolis, Ind
We thank Drs Leonidas and Berdon for their thoughtful consideration of our research and its application to urinary tract infection imaging in children. We agree that estimates of the incidence of postpyelonephritic hypertension and renal failure derived from older studies that used excretory urography to detect scarring are probably overestimates of the true sequelae of scarring. Scars visible at excretory urography are likely larger than those seen at MR imaging and scintigraphy, and it is reasonable to expect that the larger the parenchymal loss the more likely that sequelae of hypertension and renal insufficiency will occur. The degree to which these conditions are overestimated is difficult to quantify, however. Studies that include long-term follow-up of children who have imaging-documented pyelonephritis would be helpful to (a) evaluate the true incidence of scarring, (b) determine whether we can predict which children will be at greatest risk for scarring, and (c) evaluate the true incidence of the sequelae of hypertension and renal insufficiency.
The purpose of our study was to compare MR with the accepted standard in childhood pyelonephritis imaging, 99mTc DMSA, to determine whether MR is as good as 99mTc DMSA in this capacity (1). MR imaging compared favorably with 99mTc DMSA scintigraphy, and, in our opinion, it may be used as an alternative to 99mTc DMSA imaging for the detection of pyelonephritis. An excellent point raised by Drs Leonidas and Berdon is that all children with urinary tract infection (and even febrile urinary tract infection) do not necessarily need MR or any other imaging to document pyelonephritis, and this is certainly true if clinical management is not altered. Many hospitals treat children with pyelonephritis by means of intravenous therapy. Also, in some centers, children with imaging-documented pyelonephritis but without demonstrable vesicoureteral reflux receive suppressive antibiotic therapy to prevent future episodes of pyelonephritis (2). At our prior institution, where this study was performed, the documentation of pyelonephritis at 99mTc DMSA or MR imaging led to a change in long-term clinical management, especially in patients without demonstrable vesicoureteral reflux; however, this is not the case at all institutions. Some clinicians use prophylactic antibiotic therapy for patients with clinical pyelonephritis, even in the absence of vesicoureteral reflux, and parenchymal imaging may have little or no effect on the treatment of their patients.
Drs Leonidas and Berdon suggest the judicious use of imaging studies for pyelonephritis, and we agree. There are select groups of patients in whom we believe parenchymal imaging would be particularly helpful: (a) patients with fever of unknown origin and inconclusive evidence of urinary tract infection, (b) patients with chronic bacteriuria (such as those who undergo intermittent catheterization) who have pain or fever, which could be a urinary tract infection or infection from another source, (c) children treated with suppressive antibiotics in whom febrile urinary tract infection develops, and (d) children with recurrent clinical pyelonephritis without demonstrable VUR. For now, the decision of which patients should undergo imaging may best be made on a case-by-case basis.
Again, we thank Drs Leonidas and Berdon for their discussion of urinary tract infection in children. Considerable uncertainty and controversy surround the entity of pediatric urinary tract infection, and there is much yet to be learned about this common childhood infection and its long-term effects on health.
References
This article has been cited by other articles:
|
|
 |
|
  W. H. McAlister, S. K. Fernbach, K. A. Feinstein, and M. B. Schmidt Invited Commentary • Authors' Response RadioGraphics, January 1, 2000; 20(1): 168 - 171. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
