J Clin Ultrasound 14595-600.

October 1986

Increased Renal Cortical Echogenicity

in Pediatric Renal Disease:
Histopathologic Correlations
A. Norman Brenbridge, MD,* Robert L. Chevalier, MD,t and
Donald L. Kaiser, PhD$

Abstract: Retrospective comparisonsbetween sonographicrenal cortical echogenicity

and the results of renal biopsies were made for 65 pediatric patients ranging in age
from neonate to 18 years. There was a positive correlation between an increase in renal
cortical echoes and interstitial infiltration as well as with glomerular obsolescence,
tubular atrophy, and vascular changes. Since the sonographic changes were hetero-
geneous in origin, they are not specific. The sonographic findings also correlated pos-
itively with the clinical severity of disease. However, the heterogeneous origins of the
sonographic finding and the absence of strong correlations indicate that gray-scale
sonography cannot act as a prognostic index of the type or severity of disease in pediatric
patients. Indexing Words: Pediatric renal disease * Cortical echogenicity * Prognostic
indices

Ultrasonography is now well established as a use- evaluations followed within 24 h by sonographi-

ful tool in the delineation of abdominal anatomy.
It is especially valued as a method for imaging
the kidneys in pediatric patients since it utilizes
no ionizing radiation, is noninvasive, and is tol-
erant of some patient motion. Considerable effort
has been expended in characterizing the normal
echographic patterns of children’s kidneys and in
determining what constitutes a deviation from the
normal. However, the question still remains what
significance to place in any deviations from the
norm. We addressed the question of the relation-
ship between sonographic findings and the clinical
and pathologic parameters of pediatric renal dis-
ease by correlating the sonographic findings with
the pathologic results of pediatric renal biopsies
performed during a 5-year period.
MATERIALS AND METHODS
Between March 1979 and March 1984, 77 pedi-
atric patients with renal disease had sonographic
From the *Department of Radiology, tDepartment of Pediat-
rics, and $Department of Internal Medicine, Division of Bio-
statistics, University of Virginia Medical Center, Charlottes-
ville, Virginia. For reprints contact A. Norman Brenbridge,
MD, Department of Radiology, Box 170, University of Virginia
Medical Center, Charlottesville, Virginia 22908.
0 1986 by John Wiley & Sons, Inc.
0091-2751/86/080595-06 $04.00
cally guided biopsies a t the University of Virginia.
The data from 65 patients who had both techni-
cally adequate sonograms and adequate biopsy
specimens were analyzed retrospectively. The
sonographic and pathologic findings were evalu-
ated independently.
The patients ranged in age from 1 week to 18
years at the time of biopsy (mean age 9.86 [ & 5.98
SDI years); there were 32 females and 33 males.
All had clinical renal disease based on clinical and
laboratory findings. None had clinical or labora-
tory evidence of hepatic, splenic, or hematologic
disease.
The clinical-pathologic diagnoses are listed in
Table 1. The patients were individually scored 0
to 4 for increasing clinical severity of disease based
on decreasing creatinine clearance (Table 2).
All sonograms were obtained with commercial
units utilizing 3.5-MHz, S-MHz, and 6-MHz trans-
ducers. The TGC curves had been adjusted for each
transducer and patient; gains had been altered to
optimal levels according to the individual’s size
and body habitus. All patients had dynamic (real-
time) scans, and most also had static (contact)scans.
The left and right kidneys were evaluated inde-
pendently for the intensity of cortical echoes. In
two patients the left kidney was sufficiently well
595
596 BRENBRIDGE ET AL.

TABLE 1 TABLE 3
Distribution of Diagnoses Renal Cortical Echogenicity
Clinical-Pathologic Diagnosis Frequency Percentage Frequency Percentage
Negative 5 7.7 Grade 1: echogenicity less than 6 4.7
Lipoid nephrosislminimalchange 15 23.1 liver/spleen
Lupus nephritis 5 7.7 Grade 2: echogenicity equal to 73 57.0
Glomerulonephritis, idiopathic 15 23.1 liverkpleen
Glomerulonephritis, familial and 3 4.6 Grade 3: more echogenic than 37 28.9
hereditary liverlspleen
End-stage renal disease 1 1.5 Grade 4: echogenicity equal to central 12 9.4
Interstitial nephritis 1 1.5 renal sinus or portal vein
Henoch-Schonlein disease 4 6.2 Total 128 100
Berger's disease 1 1.5
Senior's syndrome 1 1.5
Wegener's syndrome 3 4.6
Sickle cell glomerulopathy 1 1.5
Diabetic glomerulosclerosis 4 6.2 be slightly more echogenic than the liver. None
Congenital nephrotic syndrome 2 3.1 of our four infants fell into these two categories.
Nephrocalcinosis 2 3.1
Thrombosis, arterial 1 1.5
A score of 3 was assigned to echogenicity signif-
Thrombosis, venous 1 1.5 icantly greater than that of the liver (three term
Total 65 100.0 neonates 1, 2, and 3 wk of age, respectively). A
score of 4 was granted to intense cortical echo-
genicity approximately equal to that surrounding
seen to permit successful biopsy but was techni- the portal vein (one term infant of 2 months) (Ta-
cally inadequate for echographic analysis. These ble 3).
two kidneys were eliminated from the study, but In all cases, sonographically guided renal biop
the right kidneys were retained. sies were obtained from the lower pole of the left
Using the well-known relationships between kidney using either a Vim-Silverman needle with
renal cortical echogenicity and that of adjacent Franklin modification or disposable Tru-Cut
structures (the liver, spleen, and renal sinus), the (Travenol Laboratories) biopsy needle. Written
echogenicity of the renal cortex was scored on a
1to 4 basis.14 For the 62 children over 6 months
of age, a score of 1 was given to normal cortical
echogenicity less than that of the adjacent liver
or spleen. A score of 2 was given to those with
modestly increased echogenicity equal to that of
the liver or spleen. A score of 3 was utilized for
those cortices more echogenic than liver or spleen,
and 4 was given those with most marked echo-
genicity, equaling that of the central renal sinus
(Table 3) (Figs. 1-4).
For our four infants below 6 months of age, in
whom it is recognized that renal cortical echoes
may normally be as echogenic as the adjacent liver
or spleen and in whom renal sinus echoes are often
poorly defined, the scoring of the echogenicity was
a l t e ~ - e d . ~A- ~score of 1 was proposed for echoes
less than or equal to that of the liver or spleen,
and a score of 2 for those subjectively judged to

TABLE 2
Clinical Severity of Diseases Based on Cteatinine Clearance
__ -. -- FIGURE 1. Parasagittal sonogram through the right kidney of a 13-
0 = Normal or in remission (creatinine clearance > 90') year-old male with a negative renal biopsy. The renal cortical echoes
1= Mild (creatinine clearance > 75) (largearrow) are less intense than those of the liver (LI. This is normal,
2 = Moderate (creatinine clearance = 50-75) grade 1. Within the kidney, the intensely reflective central complex is
3 = Severe (creatinine clearance = 10-50) the renal sinus (S). The poorly echoic spaces are the renal pyramids
4 = End-stage renal disease (creatinine clearance < 10)
(small arrow). At the bases of the pyramids are linear echoes, which
'Creatinine clearance given in mllmin per 1.73 m*. are the arcuate vessels demarcating the corticomedullary junction.

JOURNAL OF CLINICAL ULTRASOUND

 RENAL CORTICAL ECHOGENICITY
ffiURE 2. Parasagittal sonogram through the right kidney (crosses)
of an 11-year-old male with lipoid nephrosis. The renal cortex is is-
oechoic with the adjacent liver. This is grade 2.
informed consent was obtained from each child's
parent(s). The renal biopsies were evaluated by
pathologists using routine histopathologic cri-
teria.
A single examiner (ANB)classified the sono-
graphic findings with the knowledge that all pa-
tients had clinical renal disease necessitating bi-
opsy but without prior knowledge of the clinical
information or biopsy results. Two examiners (ANB
and RC)reviewed the pathologic reports and clas-
sified the results without reference to the sono-
graphic findings. Four histologic characteristics
were evaluated: glomerular alterations, intersti-
tial disease, tubular atrophy, and vascular cel-
lular infiltration and sclerosis. These were then
evaluated 0 (normal) to 4 dependent on the in-
creasing severity of disease (Table 4). Glomerular
changes were further divided into four subgroups
based on the type of glomerular involvement,
whether focal proliferative, diffuse proliferative,
focal nonproliferative, or diffuse nonproliferative
(Table 5).
Renal cortical echogenicity was correlated with
the estimated severity of disease, the type of dis-
ease, and individual pathologic criteria, including
the type of glomerular disease and the extent of
glomerular, interstitial, tubular, and vascular in-
volvement. Age and sex relationships were also
evaluated. Pearson correlation coefficients were
used to examine relationships among interval in-
dices. This method was derived using continuous
interval data, which are assumed to be drawn from
VOL. 14, NO. 8, OCTOBER 1986
597
FIGURE 3. Parasagittal sonogram through the right kidney of a 10-
year-old female with systemic lupus erythematosus. The renal cortical
echoes are more intense than those of the adjacent liver but less than
those of the central renal sinus. This is grade 3.
normally distributed populations. The Pearson
technique is very robust; the probabilities derived
from its application retain substantial accuracy
even when some of the assumptions of their de-
rivation have been violated. It is also one of the
most available programs for computer analysis.
Therefore, this correlation coefficient is widely ap-
plied in studies such as that reported here, al-
though it was not originally designed for discon-
tinuous data.
Differences between the sexes for individual in-
dices were examined using a t statistic. Because
many relationships among indices were exam-
ined, a more conservative a level (P< 0.01)was
used to consider the significance of individual re-
sults.
RESULTS
We found no significant age relationships, and this
factor was eliminated from further considera-
tions. There were slight differences between the
sexes for various criteria. The differences were not
statistically significant and were likely a result
of the smaller sample sizes when the data were
analyzed by sex. However, there may be some pro-
598 BRENBRIDGE ET AL.

flGURE 4. Parasagittal sonogram through the right kidney of a 16-year-old male with Wegener's disease.
The renal cortical echoes are not onlv more intense than those of the liver but eaual to those of the central
renal sinus. This is grade 4.

pensity for certain diseases to occur in one sex or
the other. Sex was eliminated from further anal-
ysis.
Renal cortical echogenicity (N = 128) corre-
lated most strongly with the degree of interstitial
infiltration (r = 0.43883, P = 0.0003). There was
a weaker positive correlation between echogen-
icity and the severity of glomerular disease (r =
0.32615, P = 0.0080). No specific differences were
found among the four subgroups of glomerular
disease.
The correlation (r = 0.33062,P = 0.0071)of
vascular infiltration and sclerosis with echogen-
icity was similar to that for glomerular disease.
Tubular atrophy had the weakest correlation with
echogenicity (r = 0.31583, P = 0.0104) of those
four parameters.
Increasing cortical echogenicity had a positive
correlation ( r = 0.45802, P = 0.001) with clinical
severity.
DISCUSSION
Renal sonography is usually performed in chil-
dren who have palpable abdominal masses, oli-
guria, or clinical or laboratory evidence of renal
abnormalities. Sonographically, kidneys are ex-
amined for size, position, and contour, and note is
made of the relative echogenicity and pattern of
the cortex and medulla. Urosurgical conditions
are usually distinguished from medical conditions
with ultrasonography, and the patient's evalua-
tion is directed appropriately.
Medical renal disease may affect any or all of
the four constituents of the kidney, the glo-
merulus, the tubules, the interstitium, and the

TABLE 4
Pathologic Severity of Disease
Grade Glomerular Interstitial Tubular Vascular
0 No change No change No change No change
1 Minimal change Minimal cellular Minimal Slight intra- or perivascular
infiltration or fibrosis cellular infiltrate
2 < 50% glomerular Moderate cellular Focal atrophy Prominent intra- or
obsolescence infiltration or fibrosis perivascular cellular
infiltrate
3 50-90% glomerular Marked cellular Marked atrophy Marked intra- or perivascular
obsolescence infiltration or fibrosis cellular infiltrate
4 > 90% glomerular Severe cellular infiltration Severe atrophy Vascular sclerosis
obsolescence and fibrosis

JOURNAL OF CLINICAL ULTRASOUND

 RENAL CORTICAL ECHOGENICITY
TABLE 5
Distribution of Glomerular Pathologies
- ~ ~
Glomerular Pathology Frequency Percentage
-~
Normal 15
Focal proliferative 9 13.8
Diffuse proliferative 25
Focal nonproliferative 10
Diffuse nonproliferative 6 9.2
Total 65 100
- _ _ -~ -~
vessels. The nature and extent of the involvement
vary with the type and severity of disease, and
the primary effect on any constituent(s) of the kid-
ney may affect the other(s). Diseases are catego-
rized according to the nature and distribution of
renal involvement. Carried to their ultimate con-
clusion, in end-stage renal disease, all constitu-
ents are so severely altered that the original dis-
ease may be unrecognizable pathologically.
Prognosis and therapeutic plans are based on
the diagnosis and severity of disease. Although
clinical and laboratory data may be suggestive of
the identity of the disease process and its severity,
renal biopsy is, at present, the only accurate method
of diagnosis. This is also true for the monitoring
of the progression or regression of the disease.
Although percutaneous biopsy is not without in-
herent short- and long-term risks, these are gen-
erally acceptable. Still, one might hope for a non-
invasive method of diagnosis. There are many
ongoing efforts nationwide to characterize tissues
sonographically, but no reports to date of un-
qualified success.
Although tissue typing is not yet possible, gray-
scale ultrasonography is still a useful modality.
This is especially true in the pediatric population
in whom there is a great desire to avoid invasive
or ionizing procedures. Real-time sonography is
also tolerant of the almost inevitable movement
encountered in young children.
Among various sonographic factors, particular
attention has been paid to renal cortical echogen-
icity. In normal kidneys, the renal cortex is less
echogenic than is the adjacent liver or spleen and
considerably less echogenic than the renal si-
nus.1:2 The exception is in the neonatal and early
infantile period; under 6 months of age, renal cor-
tical echogenicity may be equal to that of the liver?
(All four of our neonates had hyperechogenic cor-
tices, and all four had profoundly diseased kidneys
pathologically.) Certain patterns have been sug-
gested, especially for those with polycystic kidney
disease (both adult and infantile forms) a s well as
end-stage renal disease (ESRD) of any etiology.
However, except for ESRD, increased echogenic-
599
ity does not seem to differentiate acute, subacute,
or chronic phases of any disease.’
The mechanism of increased echogenicity is un-
certain, although a variety of possibilities have
23.1
been postulated, including altered perfusion, cell
38.5 infiltration, and deposition of connective tissue,
15.4 calcium, and fat. In a 1982 study of 109 biopsies
on patients betwen ages 8 years and 78 years,
- Hricak et al.3 found a positive correlation between
increased cortical echogenicity and glomerular
changes. In 1981, Rosenfield and Siege12 had al-
ready noted a positive correlation between cortical
echogenicity and interstitial changes in 25 pa-
tients but did not find a relationship to the glo-
merular lesions.
We sought to find what relationships, if any,
held true for pediatric patients. It is well accepted
that the disease of children may differ from those
of adults, as may the manifestations of disease.
Their sonographic patterns may differ in various
age groups, and, as a consequence of these sup-
positions, so may their sonograpbic responses to
Although renal cortical echogenicity correlated
most closely with interstitial infiltration, corre-
lations with glomerular, vascular, and tubular af-
flictions were similar. (The respective p2 values
for interstitial, glomerular, vascular, and tubular
changes are 19.2%,10.6%,10.9%,and 10.0%.)This
leads one to conclude that hyperechogenicity of
the renal cortex is a multifactorial response re-
lated to multiple alterations in the structural in-
tegrity of the tissue.
Five of the biopsies were interpreted as without
histologic change by light microscopy. Three of
these children bore the clinical diagnosis of sys-
temic lupus erythematosus and two of interstitial
nephritis. Of these five, three showed no signifi-
cant electron microscopic abnormalities. Two
showed slight basement membrane abnormalities
on electron microscopic examination (both were
lupus patients). All five patients had laboratory
evidence of prior renal dysfunction, but four were
in clinical remission with grade 0 clinical severity
(normal creatinine clearance), and one (with in-
terstitial nephritis) had only mild dysfunction
(grade 1 severity, creatinine clearance over 75%).
The cortical echoes of all 10 kidneys were grade
2, slightly increased. If these kidneys were ac-
tually affected in some way, either our biopsy
specimens missed the abnormalities or changes
are not detectable by routine histologic procedures
(light, electron, and immunofluorescence micros-
copy).
It has also been tempting to speculate that the
degree of change of renal cortical echogenicity may
600 BRENBRIDGE ET AL.
act a s a prognostic indicator of the severity of pe-
diatric disease. By grading the severity of echo-
genicity, one might then have an index of the de-
gree of pathologic involvement and, possibly, its
etiology. Unfortunately, this does not appear to
be the case. There are many variables changing
to different degrees in various disease states. No
one index has an overwhelmingly positive rela-
tionship. Comparison of increasing cortical echo-
genicity with overall clinical severity does have
a positive correlation coefficient ( r = 0.45802,
P = 0.0001). However, an ? of 20.9%,although
a positive correlation, is far from a strong one.
Therefore, neither this category nor any of the
other lesser positive correlations obtained be-
tween cortical echogenicity and types of cortical
alteration may be considered definitive.
CONCLUSION
Certain patterns of sonographic renal cortical
changes have been reported in specific congenital
diseases and end-stage renal disease. In medical
renal diseases occurring in pediatric patients, there
is a positive correlation between the severity of
disease and increased cortical echogenicity. How-
ever, the correlation appears to be nonspecific and
is not strong enough to be considered a prognostic
index of the severity of disease, nor is it indicative
of the type of pathologic involvement.
REFERENCES
1. Rosenfield AT, Taylor KJW, Crade M, et al: Anat-
omy and pathology of the kidney by gray scale ul-
trasound. Radiology 128:737-744, 1978.
2. Rosenfield AT, Siege1 NJ: Renal parenchymal dis-
ease: Histopathologic-sonographic correlation. A m
J Roentgen01 137:793-798,1981.
3. Hricak H, Cruz C, Romanski R Renal parenchymal
disease: Sonographic-histologic correlation. Radiol-
ogy 144:141-147, 1982.
4. Haller JO, Berdon WE, Friedman A P Increased renal
cortical echogenicity: A normal finding in neonates
and infants. Radiology 142:173-174, 1982.
5. Hricak H, Slovis TL, Callen CW, et al: Neonatal
kidneys: Sonographic anatomic correlation. Radiol-
ogy 1471699-702, 1983.
6. Hayden CK, Santa-Cruz FR, Amparo EG, et al: Ul-
trasonographic evaluation of the renal parenchyma
in infancy and childhood. Radiology 152:413-417,
1984.
7. Krensky AM, Reddish JM,Teele R L Causes of in-
creased renal echogenicity in pediatrics patients. Pe-
diatrics 722340-846, 1983.
8. Babcock DS, Slovis TL, Han BK, et al: Renal trans-
plants in children: Long-term follow-up using so-
nography. Radiology 156:165-167, 1985.
JOURNAL OF CLINICAL ULTRASOUND