Glomerular and Tubular Dysfunction in
Children with Congenital Cyanotic Heart
Disease: Effect of Palliative Surgery
HESHAM AWAD, MD; IBRAHIM EL-SAFTY, PHD; MOUSTAFA ABDEL-GAWAD, MD;
SALWA EL-SAID, MSC
ABSTRACT: Background: Nephropathy has long been
recognized as a potential complication of congenital
cyanotic heart disease (CCHD). The present study was
undertaken to investigate some aspects of glomerular
function by measuring urinary total protein, microalbu-
min, and tubular function by assessing urinary ␣-1-
microglobulin. The structural integrity of the renal prox-
imal tubules was also studied by measuring urinary
activities of the brush-border enzyme leucine-amino-
peptidase and the lysosomal enzyme N-acetyl-␤-D-glu-
cosaminidase. The levels of hematocrit (Hct) and oxygen
saturation were also investigated as predisposing factors
for renal impairment in CCHD. Methods: These investi-
gations were done by recruiting 86 children who were
grouped as follows: the control group (G1) consisted of
14 children (aged 4 –12 years); the other 72 children
with CCHD were divided according to age (ie, duration
of cyanosis) into 4 equal groups, each containing 18
patients: G2 (age ⬍1 year), G3 (age ⱖ1 year and ⬍5
years), G4 (age ⱖ5 years and ⬍10 years), and G5 (age
ⱖ10 years). In addition, 10 of the 72 patients underwent
a palliative surgery and were included as G6 (regard-
less of age: 2 from G3, 4 from G4, and 4 from G5) to
study the effect of the palliative surgery on the above-
N ephropathy has long been recognized as a po-
tential complication in patients with long-
standing CCHD1. The incidence of renal abnormal-
ities increases with the degree of cyanosis and may
also increase with increasing duration of cyanosis2.
Abnormalities of renal structure and function oc-
cur in some patients with CCHD. The principle
renal structural lesions have been in the glomeru-
lus,3 and tubular changes are rare.4 Abnormal renal
From the Departments of Pediatrics (HA, SE-S) and Cardiology
(MA-G), Faculty of Medicine, and Section of Biochemistry, Depart-
ment of Chemistry, Faculty of Education (IE-S), Ain-Shams Uni-
versity, Cairo, Egypt.
Submitted July 26, 2002; accepted December 10, 2002.
Correspondence: Ibrahim El-Safty, Ph.D., Department of Bio-
chemistry, Faculty of Science, Makerere University, PO Box 7062,
Kampala, Uganda.
110
mentioned parameters. Results: Results of the present
work showed that with increasing duration of cyano-
sis (ie, on going from G2 to G5) among the studied
children with CCHD, there was a significant elevation
in the urinary excretion of the investigated functional
and structural parameters of the glomeruli and proxi-
mal tubules compared with the control children. The
data also showed a significant increase in Hct,
whereas oxygen saturation was significantly de-
creased. Results of G6 after the palliative surgery
demonstrated a significant decrease in the urinary
excretion of the investigated parameters of the kidney,
with a significant decrease in Hct and increase in
oxygen saturation levels, compared with the results of
the patients of this group before the palliative surgery.
Conclusions: These results suggest impairment of both
glomerular and tubular functions as well as structure of
the proximal tubules among children with CCHD and that
the palliative surgery has significantly improved this
impairment. KEY INDEXING TERMS: Congenital cyanotic
heart disease (CCHD); Children; Palliative surgery; ␣-1-
microglobulin; Leucine-aminopeptidase; N-acetyl-␤-D-
glucosaminidase; Hematocrit; Oxygen saturation. [Am J
Med Sci 2003;325(3):110–114.]
functions occur in some patients with CCHD.2 Clin-
ical studies have reported these functional abnor-
malities, mostly without discriminating between
glomerular and tubular dysfunction.4 Findings in-
cluded normal5 or decreased6 glomerular filtration
rate (GFR) and decreased renal plasma flow2 as well
as both glomerular and tubular proteinuria7 were
reported.
Because of the great reserve capacity of the kid-
ney, tests used for assessing the functional integrity
of the kidney (such as serum creatinine or urea and
GFR) are not sufficiently sensitive because these
measurements can be in the normal range despite
considerable impairment of renal function, and they
become significantly altered only after considerable
damage to the kidney has occurred with major func-
tional impairment8. Various urinary parameters of
March 2003 Volume 325 Number 3
 Awad et al

Table 1. Laboratory Reference Values for the Investigated the kidney functional and structural integrity
Urinary Parameters among 14 Healthy Children (Aged 4 –12
Years)
among such patients was investigated.

Parameter Range (mean ⫾ SD) Subjects and Methods

U.TP/U.Cr mg/g Cr 29.1–80.95 (49.96 ⫾ 14.59)
U.Malb/U.Cr ␮g/mg Cr 1.86–5.75 (3.47 ⫾ 1.10)
U.␣1M/U.Cr ␮g/mg Cr 2.35–11.67 (6.93 ⫾ 4.14)
U.LAP/U.Cr U/mg Cr 1.44–15.86 (9.74 ⫾ 4.57)
U.NAG/U.Cr nmol/mg Cr 0.99–24.73 (8.94 ⫾ 7.63)
TP, total protein; U.Cr, urinary creatinine.
the kidney, such as microalbumin (Malb) and ␣-1-
microglobulin (␣1M) have proved useful to assess
functional integrity of glomerulus and proximal tu-
bules, respectively, whereas urinary kidney-specific
enzymes such as brush-border leucine-aminopepti-
dase (LAP) and lysosomal N-acetyl-␤-D-glucosamini-
dase (NAG) are indicators for structural integrity of
proximal tubules.9
The purpose of this study was to evaluate some
aspects of glomerular and proximal tubular func-
tions as well as structural integrity of the proximal
tubules in children with CCHD. The study also
aimed to investigate the role of both hematocrit and
oxygen saturation as predisposing factors of renal
impairment among such patients. The effect of pal-
liative surgery on the above-mentioned factors and
Subjects. Fourteen healthy children aged 4 to 12 years served
as control group (G1). Patients of the present study were recruited
over a period of 6 months (June–December 1999) from those
attending Ain-Shams University Childrens’ Hospital and Insti-
tute of Cardiology, Cairo, Egypt. The study was approved by the
hospital review board. Patients were subjected to full medical
history, clinical examination, and echocardiology. Patients with
coarctation of the aorta, associated renovascular or structural
malformations, urinary tract infection, or receiving antibiotics of
aminoglycoside family were excluded. After application of these
exclusion criteria, 72 patient children suffering from CCHD were
included in the present study. They were divided into 4 equal
groups (each obtained 18 cases) according to age: G2 (⬍1 year), G3
(ⱖ1 year and ⬍5 years), G4 (ⱖ5 years and ⬍10 years) and G5 (ⱖ10
years). In addition, 10 of the 72 cases (2 from G3, 4 from G4, and
4 from G5) underwent palliative surgery and were investigated as
G6.
Sampling and Methods. Parents of each child (patient and
control) participating in the investigation gave their oral permis-
sion to get the urine sample and perform the renal function tests.
Morning urine sample was suggested as the best sample for
detecting early kidney abnormalities.10 A random morning urine
sample was collected at home between 8 and 9 AM from each
participant child (control and case subjects) in a sterilized con-
tainer and centrifuged at 4500 rpm for 5 minutes, and the top 5
mL of the supernatant were stored frozen at ⫺20°C in aliquots
without preservatives until analyzed within 2 weeks for the
assessment of:
1. Glomerular function by estimation of total protein by dye-

Table 2. Values (mean ⫾ SD) of the Investigated Parameters among the Control Group and the Different Groups of CCHD
Children

Control Different Groups of CCHD Children

G1 (n ⫽ 14) G2 (n ⫽ 18) G3 (n ⫽ 18) G4 (n ⫽ 18) G5 (n ⫽ 18)

Parameters Age 4–12 years Age ⬍1 years Age ⱖ1 year to ⬍5 years Age ⱖ5 years to ⬍10 years Age ⱖ10 years

Hct 0.35.11 ⫾ 0.274 0.3833 ⫾ 0.327* 0.4433 ⫾ 0.512*a 0.5505 ⫾ 0.445*bd 0.6235 ⫾ 0.751*cef
Oxygen saturation 95.21 ⫾ 3.11 66.22 ⫾ 6.83* 61.27 ⫾ 6.47* 61.83 ⫾ 7.53* 59.41 ⫾ 7.70*c
(%)
Glomerular function
U.TP/ 49.96 ⫾ 14.59 91.11 ⫾ 112.92* 80.95 ⫾ 93.18* 133.76 ⫾ 79.37*bd 189.30 ⫾ 95.34*ce
U.Cr (mg/g)
U.Malb/ 3.47 ⫾ 1.10 2.62 ⫾ 1.73* 14.49 ⫾ 17.74*a 73.66 ⫾ 70.74*bd 224.42 ⫾ 208.03*cef
U.Cr (␮g/mg)
Tubular function
U.␣1M/ 6.93 ⫾ 4.14 22.34 ⫾ 20.59* 36.91 ⫾ 48.01* 55.33 ⫾ 19.97*bd 83.12 ⫾ 82.63*cef
U.Cr (␮g/mg)
Tubular structure
U.LAP/ 9.74 ⫾ 4.57 12.40 ⫾ 18.31* 35.99 ⫾ 38.45*a 45.25 ⫾ 44.55*b 103.47 ⫾ 157.06*c
U.Cr (U/mg)
U.NAG/ 8.94 ⫾ 7.63 31.68 ⫾ 26.11* 75.56 ⫾ 84.58*a 201.73 ⫾ 149.86*bd 244.77 ⫾ 101.96*ce
U.Cr (nmol/mg)

*P ⬍ 0.05 compared with control group (G1).

a
G2:3, P ⬍ 0.05.
b
G2:4, P ⬍ 0.05.
c
G2:5, P ⬍ 0.05.
d
G3:4, P ⬍ 0.05.
e
G3:5, P ⬍ 0.05.
f
G4:5, P ⬍ 0.05.

THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES 111

Palliative Surgery in Congenital Cyanotic Heart Disease

Table 3. Percentage of CCHD Patients with Abnormal Level of Table 5. Correlation Coefficient (r) between Measured Urinary
the Measured Parameters in Each Investigated Group Glomerular and Tubular Markers among CCHD Children as 1
Group of Patients
Parameter G2 G3 G4 G5
Urinary Tubular Markers
Hct 11 27 44 83 Urinary Glomerular
U.TP/U.Cr 27 27 66 88 Markers U.␣1M/U.Cr U.LAP/U.Cr U.NAG/U.Cr
U.Malb/U.Cr 0 72 100 100
U.␣1M/U.Cr 73 89 95 100 U.TP/U.Cr 0.55* 0.26* 0.29*
U.LAP/U.Cr 11 61 61 70 U.Malb/U.Cr 0.53* 0.44* 0.62*
U.NAG/U.Cr 44 88 100 100
* P ⬍ 0.05.

binding method11 and microalbumin (U.Malb) by enzyme-linked
immunosorbent assay (ELISA) technique using a kit from OR-
Gentec-Germany (Mainz, Germany).
2. Tubular function by measuring ␣1-microglobulin (U.␣1M) by
ELISA technique using a kit from ORGentec-Germany.
3. Tubular structural integrity by estimation of leucine-
aminopeptidase (U.LAP)12 and N-acetyl-␤-D-glucosaminidase
(U.NAG).13
4. Urine creatinine (U.Cr) concentration by Jaffe kinetic
method.14
In addition, oxyhemoglobin saturation was assessed using
pulse oximetry, and Hct was estimated using microcapillary tube
centrifugation after obtaining a blood sample by pin-prick
method.
Spot urine measurements were used because it has been shown
that urinary protein/creatinine ratio15 and albumin/creatinine
ratio16 as well as dividing the urinary enzyme activity by urinary
creatinine concentration17 in a random urine sample correlate
with 24-hour urinary excretion and eliminate variations caused
by changing rates of urine output and provide a measure inde-
pendent of urine concentration.
Statistical Analysis. Results were tabulated as mean ⫾ SD
and analyzed using paired t test for parametric data, whereas the
nonparametric data were analyzed using the Mann-Whitney
U–Wilcoxon Rank Sum W test and sign test. Spearman’s Corre-
lation Coefficients were used and level of significance was con-
sidered at P ⬍ 0.05.
Results
Data from the present study (Table 2) showed,
compared with control subjects (G1), a significant
increase in Hct and urinary parameters of the glo-
meruli (except U.Malb/U.Cr in G2, which showed
significant decrease in the mean value) and tubules
with increasing duration of cyanosis (on going from
G2 to G5), whereas oxygen saturation was signifi-
cantly decreased. As duration of cyanosis increased,
the percentage of CCHD patients with abnormal
levels of each investigated parameter increased (Ta-
ble 3). There was a significant correlation between
age and Hct, and each was significantly correlated
with urinary parameters of the glomeruli and the
tubules (Table 4).
Both TP and Malb (glomerular urinary markers)
were significantly correlated with both ␣1M and
LAP as well as NAG (proximal tubular urinary
markers) (Table 5). Data for the effect of palliative
surgery (Table 6) showed that the levels of Hct and
urinary parameters of glomeruli and tubules were
significantly decreased, whereas oxygen saturation
was significantly increased after the surgery com-
pared with levels before the surgery.
Discussion
Results for Hct in the present study (Tables 2, 3,
and 4) are in agreement with the findings of others
that Hct was significantly elevated among CCHD
patients4,6,7 and that Hct was correlated with the
age of patients (ie, duration of cyanosis).7,18 Data of
oxygen saturation in the present work (Table 4)
confirms the report of others4 that oxygen satura-
tion at rest was related neither to Hct elevation nor
to renal injury (both at the glomerular and proximal
tubular levels).
Results of the glomerular function in the present
investigation (Tables 2 and 3) support the work of
others4,6,7 who reported increased proteinuria and
microalbuminuria among CCHD patients, indicat-
ing glomerular dysfunction. This elevation of pro-
teinuria and microalbuminuria may be attributable
to 1 of 2 possibilities or to both. First, renal histo-

Table 4. Correlation Coefficient (r) between the Investigated Predisposing Factors for Renal Impairment and Urinary Parameters of
Kidney Function among CCHD Children as 1 Group of Patients

Predisposing Factors for Urinary Parameters for

Renal Impairment Glomerular Function Urinary Parameters for Tubular Integrity

Age Hct Oxygen-Saturation U.TP/U.Cr U.Malb/U.Cr U.␣1M/U.Cr U.LAP/U.Cr U.NAG/U.Cr

Age 1 0.86* 0.13 0.38* 0.57* 0.38* 0.33* 0.63*

Hct 1 0.21 0.42* 0.53* 0.24* 0.24* 0.55*
Oxygen saturation 1 0.22 0.14 0.20 0.20 0.15

* P ⬍ 0.05.

112 March 2003 Volume 325 Number 3


Table 6. Effect of Palliative Surgery on the Investigated
Parameters among CCHD Children
G6 (Before) G6 (After)
Parameters (Mean ⫾ SD) (Mean ⫾ SD)
Hct 0.5160 ⫾ 0.105 0.4370 ⫾ 0.764
Oxygen saturation (%) 63.90 ⫾ 5.56 92.60 ⫾ 1.57
Glomerular function
U.TP/U.Cr (mg/g) 194.00 ⫾ 93.69 47.05 ⫾ 40.70
U.Malb/U.Cr (␮g/mg) 147.90 ⫾ 220.70 27.80 ⫾ 27.48
Tubular function
U.␣1M/U.Cr (␮g/mg) 46.30 ⫾ 18.49 27.05 ⫾ 12.90
Tubular structure
U.LAP/U.Cr (U/mg) 152.30 ⫾ 149.42 70.36 ⫾ 41.14
U.NAG/U.Cr (nmol/mg) 187.53 ⫾ 146.23 37.78 ⫾ 27.39
S, significant (P ⬍ 0.05).
logical studies revealed that glomerulomegaly is as-
sociated with CCHD.19 In glomerulomegaly, there is
a selective dilation of afferent arterioles20 that in-
creases the glomerular hydrostatic pressure,21
which in turn causes glomerular capillary dilation22
and increasing GFR.21 Thus, it would seem most
likely that the increase in the glomerular size prob-
ably transiently alters the size and/or the charge-
selective permeability of the glomerular basement
membrane to protein, allowing an increase in the
protein filtered through the glomerulus. Second, in
patients with CCHD, blood hyperviscosity causes an
overall increase in renal vascular resistance,5 which
reduces the renal blood flow21 with a rise in intra-
glomerular blood pressure5. Thus, as a greater frac-
tion of the plasma flow becomes ultrafiltrate, the
average protein concentration of the plasma remain-
ing within the capillary would rise, accounting in
part for the enhanced proteinuria seen in such cir-
cumstances. Therefore, hemodynamic changes in
the glomerulus may enhance proteinuria, both by
their effect on the properties of the glomerular wall
and by altering the concentration during protein
movement from capillary to the Bowman space.
It was observed that results of U.Malb/U.Cr in G2
showed significant decrease in the mean value than
the controlled group G1. This may be because pa-
tients in G2 are much younger than children in the
control group C1.
Data on the percentage of CCHD patients with
abnormal levels of glomerular urinary markers (Ta-
ble 3) and the correlations between the measured
parameters in the present study (Table 4) coincide
with the reports of others who found that glomerular
damage is related to both the duration of cyanosis
and the extent to which Hct is elevated4 and that
glomerular-type proteinuria (microalbuminuria) is
correlated with the patient age.6,7
Results of proximal tubular reabsorption function
in the present study (Table 2) receive support from
the work of others who reported tubular-type pro-
teinuria4 and elevated urinary ␤2-microglobulin,6
THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
Awad et al
which is similar to ␣1M in that it is filtered through
the glomeruli and reabsorbed almost totally by prox-
imal tubules, suggesting proximal tubular reabsorp-
P tion dysfunction among CCHD patients. This sug-
gestion receives support from the results of the
S present study (Tables 2 and 3) and a report of other
S researchers6 that demonstrated damage to proximal
tubular cells as evidenced by elevated urinary activ-
S
S ities of LAP and NAG. This damage resulted in
leakage of these enzymes into urine as well as ex-
S cessive loss of ␣1M through failure of reabsorption
by the damaged proximal tubular cells. Another
S
S
evidence for the damage and reabsorption dysfunc-
tion of proximal tubules is that the increased levels
of proteins filtered through the glomerulus because
of its dysfunction may exceed the reabsorption ca-
pacity of the proximal tubules, leading to increased
urinary microalbumin and total proteinuria concen-
trations (Table 2). This evidence is supported by
results of the present study, which demonstrated
that as duration of cyanosis increased, the percent-
age of CCHD patients with abnormal levels of prox-
imal tubular urinary markers was elevated (Table
3), and a significant correlation between glomerular
and tubular dysfunction was present (Table 5), sup-
porting the observations that proteinuria is related
to decreased tubular function4,6.
Among CCHD patients of G2 (duration of cyanosis
is ⬍1 year), it was observed that of 11 patients with
normal glomerular function (as evidenced by normal
urinary excretion of both total protein and mi-
croalbumin), 9 cases demonstrated elevation of ei-
ther urinary ␣1M level or urinary NAG activity,
suggesting that proximal tubules may be affected
before the glomeruli. This suggestion receives sup-
port from the work of others,6 who reported that
urinary NAG activity was elevated among some
CCHD patients without proteinuria. Furthermore,
among CCHD patients of G2, of 5 cases with normal
urinary excretion of ␣1M, 3 cases (60%) showed
elevated urinary NAG activity, suggesting that
proximal tubular damage may precede proximal tu-
bular reabsorption dysfunction.
Data on the effect of palliative surgery in the
present study (Table 6) support the finding that
acute changes in Hct are accompanied by acute and
reversible changes in renal hemodynamics,22,23 indi-
cating that the extent of Hct elevation plays an
important role in the pathogenesis of CCHD-related
kidney disease. The present results (Table 6) dem-
onstrated a significant improving effect of palliative
surgery on oxygen saturation and glomerular func-
tion (as indicated by decreasing urinary excretion of
both total protein and microalbumin) as well as
proximal tubular reabsorption function (as revealed
by a significant decrease in urinary ␣1M level) and
structural integrity of proximal tubules (as shown
by a significant decrease in urinary activities of LAP
and NAG).
113
Palliative Surgery in Congenital Cyanotic Heart Disease
In conclusion, both glomerular and proximal tu-
bular dysfunction as well as proximal tubular dam-
age occurs in CCHD patients. In addition, the integ-
rity of proximal tubules may be affected before that
of the glomeruli. The extent and duration of hypox-
emia as well as elevated Hct are important factors
predisposing CCHD patients to renal impairment
through hemodynamic changes in the glomerulus.
Finally, palliative surgery has a significant improv-
ing effect on Hct and oxygen saturation level as well
as renal glomerular and tubular integrity. There-
fore, operative correction of the cardiac malforma-
tion could improve renal deterioration in CCHD
patients.
References
1. Burke JR, Glasgow EF, McCredie DA, et al. Nephropathy
in congenital cyanotic heart disease. Clin Nephrol 1977;7:38 –
42.
2. Flanagan MF, Hourihan M, Keane JF. Incidence of renal
dysfunction in adults with cyanotic congenital heart disease.
Am J Cardiol 1991;68:403– 6.
3. Dittrich S, Kurschat K, Dahnert I, et al. Cyanotic ne-
phropathy and use of non-ionic contrast agents during car-
diac catheterization in patients with congenital cyanotic
heart disease. Cardiol Young 2000;10:8 –14.
4. Dittrich S, Haas NA, Buhrer C, et al. Renal impairment in
patients with long-standing congenital cyanotic heart dis-
ease. Acta Pediatr 1998;87:949 –54.
5. Burlet A, Drukker A, Guignard JP. Renal function in
congenital cyanotic heart disease. Nephron 1999;81:296 –300.
6. Akita H, Matsuoa S, Kuroda Y. Nephropathy in patients
with cyanotic congenital heart disease. Tokushima J Exp
Med 1993;40:47–53.
7. Krull F, Ehrich JHH, Wurster U, et al. Renal involvement
in patients with cyanotic congenital heart disease. Acta Pe-
diatr Scand 1991;80:1214 –9.
8. Rebel W, Bertsch T, Bode G, et al. Enzymuria as indicator
of renal pathomorphology. In: Jung K, Mattenheimer H,
Burchardt U, editors. Urinary enzymes in clinical and exper-
imental medicine. Berlin: Springer-Verlag; 1992. p. 43–72.
9. Mueller PW, Lash LH, Price RG, et al. Urinary biomark-
ers to detect significant effects of environmental and occupa-
114
tional exposure to nephrotoxins. I. Categories of tests for
detecting effects of nephrotoxins. Ren Fail 1997;19:505–21.
10. Zuppi C, Baroni S, Scribano D, et al. Choice of time for
urine collection for detecting early kidney abnormalities in
hypertensives. Ann Clin Biochem 1995;32:373– 8.
11. Watanabe N, Kamei S, Ohkubo A, et al. Urinary protein
as measured with a pyrogallol red-molybdata complex, man-
ually and in a Hitachi 726 automated analyzer. Clin Chem
1986;32:1551– 4.
12. Josch W, Dubach DU. Urine-arylamidase. Z Klin Chem
1967;5:59 – 63.
13. Thomas GH. ␤-D-Galactosidase in human urine: deficiency
in generalized gangliosidosis. J Lab Clin Med 1969;74:725–
31.
14. Bartels H, Bohmer M, Heieli C. Serum kreatininebestim-
mung ohne enteiweissen. Clin Chim Acta 1972;37:193–7.
15. Lemann J Jr., Dowmans BT. Proteinuria in health and
disease assessed by measuring the urinary protein/creatinine
ratio. Clin Chem 1987;33:297–9.
16. Woolerton J, Jary DR, Dunn PJ, et al. Urine albumin/
creatinine ration and clinical correlates in a diabetic popula-
tion. N Z Med J 1987;100:130 – 4.
17. Jung K. Enzyme activities in urine. How should we express
their excretion? Eur J Clin Chem Clin Biochem 1991;29:
725–9.
18. Perloff JK. Systemic complications of cyanosis in adults
with congenital heart disease. Hematologic derangements,
renal function, and urate metabolism. Cardiol Clin 1993;11:
689 –9.
19. Hagley MT, Murphy DP, Mullins D, et al. Decline in
creatinine clearance in a patient with glomerulomegaly asso-
ciated with a congenital cyanotic heart disease. Am J Kidney
Dis 1992;20:177–9.
20. Meessen H, Litton MD. Morphology of the kidney in mor-
bus caeruleus. Arch Pathol 1953;56:480 –7.
21. Guyton AR, Hall JE. Textbook of medical physiology, 9th
ed. Philadelphia: WB Saunders; 1996. p. 324 –5.
22. DeJong PE, Weening JJ, Donker AJM, et al. The effect of
phlebotomy on renal function and proteinuria in a patient
with congenital cyanotic heart disease. Nephron 1983;33:
225– 6.
23. Wilcox CS, Payne J, Harrison DW. Renal function in
patients with chronic hypoxemia and cor pulmonale following
reversal of polycythemia. Nephron 1982;30:173–7.
March 2003 Volume 325 Number 3