Key words ultrasonography - doppler - color - kidney transplantation - child
Introduction
Ultrasound is the method of choice for noninvasive monitoring of kidney transplants
at the bedside [1 ]
[2 ]. Standard protocols assess parenchymal integrity based on B-mode and flow imaging
techniques. For vascularity assessment color Doppler sonography (CDS) including duplex
ultrasound is generally the modality of choice. In adults, contrast-enhanced ultrasound
(CEUS) has become increasingly relevant for a more specific diagnosis of allograft
dysfunction [3 ]. Changes in renal vascularization and flow have been described in rejection, infection,
urinary retention, and drug toxicity [4 ]
[5 ].
Imaging of renal transplant vascularization with CDS is widely used because of the
good signal-to-noise ratio and good penetration into deeper structures. However, known
limitations of the method are the relatively low spatial and temporal resolution,
aliasing effects with high-amplitude flow, angle dependency, and blurring artifacts
[1 ]
[4 ]
[6 ]
[7 ]
[8 ]
[9 ]. Among alternatives, B-flow sonography (BFS) is a relatively new non-Doppler-based
technique for the direct visualization of blood flow and was introduced in 2000 [10 ]. So far, BFS is available on the ultrasound platform of one manufacturer. The technique
is based on the subtraction of received amplitudes of grayscale ultrasound resulting
in angiography-like overlap free flow images with very high spatial and temporal resolution
[11 ].
BFS was initially introduced for linear transducers and is now applicable for lower
frequency convex probes, also allowing the evaluation of deeper structures such as
abdominal organs. Side-by-side comparison with CDS showed that BFS is especially useful
in areas with simultaneous low and high blood flow and for the detection of small
vessels [12 ]. Preliminary studies have been published in adult patients with carotid artery stenosis
or ovarian torsion, and regarding the evaluation of vascularization in transplanted
livers and kidneys [13 ]
[14 ]
[15 ]
[16 ]. Pediatric studies using BFS were reported for fetal congenital cardiopathies, femoral
artery stenosis before catheterization in infants and anatomy of basal cerebral arteries
in newborns [17 ]
[18 ]
[19 ]. The aim of this study was to compare BFS with CDS for the assessment of kidney
transplant vascularization in children.
Patients and Methods
Patients
The study was approved by the institutional review board with a waiver of informed
consent. All pediatric patients with kidney transplantation and who received a protocol
ultrasound examination as part of their routine follow-up at our institution by the
same single sonographer with corresponding CDS and BFS images between January 2013
and January 2016 were retrospectively assessed. If multiple examinations were available
during this period, the most recent one was chosen. Of 47 consecutive cases performed
during this period, 7 cases had to be excluded because of incomplete documentation
or artifacts attributable to non-compliance. In total, 40 patients were included in
this study (mean age 11 ± 4 years, range 1–18 years; 24 male, 16 female). The mean
interval between kidney transplantation and the ultrasound examination was 1664 ± 1420
days (range 1–4820 days). Clinical data and laboratory findings of the patients were
extracted from the patient record and are summarized in [Table 1 ].
Table 1
Clinical patient data.Tab. 1 Klinische Patientendaten.
patient
current medication
current laboratory parameters
age (y)
BMI
days after TX
TX type
diagnosis leading to NTX
antihypertensives
immuno-suppressants
eGFR
urea (mg/dl)
creatinine (mg/dl)
cystatin C (mg/l)
1 m
17
19.7
2368
cad
1
1, 3, 4
1, 2, 4
53.89
15
1.25
1.08
2 m
4
17.1
793
cad
1
1, 3, 4
2, 4, 5
82.26
29
0.49
1.06
3w
8
19.4
390
cad
1
3
1, 2, 4
131.95
13
0.4
1.22
4w
8
16.6
897
cad
14
1, 3, 4
1.2
112.89
20
0.48
1.51
5 m
17
23.6
387
liv
16
1.3
1, 2, 4
31.19
34
2.25
n. a.
6 m
17
19.8
3301
cad
11
n. a.
n. a.
7.47
49
8.86
7.1
7w
4
27.3
555
liv
4
1, 3, 4
1, 2, 4
123
n. a.
n.a
0.77
8w
17
18.5
4816
liv
15
1
1, 4, 7
57.49
29
1.14
n. a.
9w
10
23.1
414
n. a.
1
1.3
1, 2, 4
68.56
19
0.95
1.22
10 m
2
15.3
445
liv
16
3
1.6
23.28
39
1.49
2.4
11w
6
16.5
201
cad
4
1.6
110.48
20
0.4
0.69
12w
12
23.2
2018
cad
1
1.2
54.23
22
1.07
1.03
13 m
10
14.7
2385
cad
9
1.4
1, 1, 2
51.81
20
1.1
n. a.
14 m
8
17.7
407
liv
16
3
1, 2, 4
90.11
14
0.55
0.99
15 m
4
16.3
1221
cad
1
5.2
49.79
39
0.9
1.6
16w
11
15.5
2695
cad
8
1.2
62.74
31
1.04
1.34
17w
12
17
3589
cad
4
1, 4, 7, 7
54.38
31
1.2
1.33
18w
17
18
1728
cad
4
1.3
1, 2, 4
46.36
24
1.47
1.48
19 m
9
18.2
790
cad
14
1, 3, 4, 5
1, 2, 4
35.12
20
1.43
1.78
20 m
11
17.6
779
liv
8
3
1, 4, 7
37.4
51
1.59
1.64
21 m
1
14.8
449
cad
9
1, 3, 4, 5
2, 4, 5
20.1
47
3.4
2.53
22 m
9
17.5
2778
n. a.
12
1, 3.
1, 4, 7
33.31
39
1.55
1.83
23 m
14
17.8
1
n. a.
12
n. a.
n. a.
8.06
85
8.97
n. a.
24 m
10
16.3
1971
cad
16
3.4
1.1, 6
43
n. a.
n. a.
1.56
25w
14
20.3
135
cad
14
1, 3, 4, 5, 6, 8
1, 2, 4
45.4
31
1.46
1.24
26 m
13
21.2
3706
cad
1
1.7
61.43
23
1.03
1.28
27 m
9
17.5
356
liv
4
1.3
1.6
49.71
20
1.11
1.34
28 m
10
15.4
90
cad
1
0
2, 4, 5
107.38
15
0.5
n. a.
29w
16
27.1
2079
liv
1
1.3
1.7
99.12
10
0.67
0.73
30 m
15
20.6
2014
cad
1
1.6
50.98
24
1.4
1.42
31w
12
19.3
2707
cad
7
3.4
2.5
59.53
21
1.11
1.29
32w
16
20.2
2372
liv
1
1
1.3
74.25
16
0.9
0.73
33 m
18
20.7
4820
cad
4
2.5
38.9
n. a.
n. a.
n. a.
34 m
15
17.3
4293
cad
9
5, 6, 7
3, 4, 6
19.84
55
3.33
2.9
35 m
1
17.1
14
cad
12
n. a.
n. a.
6.23
85
4.97
n. a.
36w
17
23.7
3234
cad
4
4
1.2
61.71
17
1.02
n. a.
37 m
7
14.7
2204
liv
7
5.7
32.24
28
1.5
1.4
38w
15
20.4
2709
cad
4
1, 3, 4
1
17.09
43
3.48
n. a.
39 m
15
15.7
454
cad
16
2, 3, 4
1, 3, 4
47.52
20
1.46
0.92
40 m
7
14
10
n. a.
n. a.
n. a.
n. a.
n. a.
n. a.
n. a.
n. a.
y = years, BMI = body mass index, TX = transplantation, NTX = kidney transplantation,
cad = cadaveric, liv = living, diagnosis leading to NTX: 1 = polycystic kidney disease;
2 = glomerulonephritis; 3 = hydronephrosis; 4 = kidney dysplasia; 5 = perinatal asphyxia;
6 = reflux nephropathy; 7 = Denys-Drash syndrome; 8 = hemolytic-uremic syndrome; 9 = CAKUT
(congenital anomalies of the kidney and urinary tract); 10 = obstructive uropathy;
11 = prune belly syndrome; 12 = oxalosis; 13 = cystinosis; 14 = nephrotic syndrome;
15 = uro-anogenital malformation; 16 = posterior urethral valve, antihypertensives:
1 = Delix (ramipril + hydrochlorothiazide); 2 = Enalapril; 3 = Amlodipine; 4 = Beloc-Zok
(metoprolol succinate + hydrochlorothiazide); 5 = Dihydralazine, immunosuppressants:
1 = tacrolimus; 2 = mycophenolate mofetil; 3 = mycophenolate; 4 = prednisolone, eGFR = estimated
glomerular filtration rate calculated as in KDIGO 2012 Clinical Practice Guideline
for the Evaluation and Management of Chronic Kidney Disease, n. a. = not available.
Ultrasound monitoring
Image optimization including the CDS, power Doppler and BFS techniques was performed
in a preliminary series of patients not included in this study. Power Doppler showed
equivalent sensitivity regarding vessel delineation but more movement artifacts than
CDS. To reduce examination time, power Doppler was consequently excluded from the
final ultrasound protocol for pediatric patients with kidney transplantation. All
examinations were performed by a single pediatric radiologist with 12 years of renal
transplant ultrasound experience, using a commercial scanner (Logiq E9, GE Medical
Systems, Milwaukee, WI, USA) with a curved transducer (C1–6) and a linear, high-resolution
transducer (ML6–15). The standard technical settings for the C1–6 transducer were
– frame rate (FR) 5, pulse repetition frequency (PRF) 1.4, mechanical index (MI) 1.2,
thermal index in soft tissue (TIs) 1.0 for CDS and – FR 16, pulse repetition interval
(PRI) 12, MI 1.2, TIs 1.1 for BFS. A high detection speckle reduction imaging (SRI
HD) strength of 2 (ranging from 0–4) was used. The setting for the CrossXBeam technique
was low, the CrossXBeam-type mean. The standard settings for the ML6–15 transducer
were – FR 11, PRF 1.5, MI 0.6, TIs 0.5 for CDS and – FR 18, PRI 10, MI 0.4, TIs 0.7
for BFS. A high detection speckle reduction imaging (SRI HD) strength of 1 was used.
The setting for the CrossXBeam technique was low, the CrossXBeam-type mean.
B-mode evaluation of the kidney in the longitudinal plane with a curved array was
followed by documentation of vascularization: (1.) CDS images during systole with
adapting of the color Doppler sensitivities from 15 to 5 cm/s; (2.) Identical BFS
image during systole without altering the probe position or device setting; (3) Flow
analysis of the interlobar and interlobular arteries in the lower, middle, and upper
part of the transplant. The renal artery and vein were evaluated in the transverse
plane. Thereafter, linear transducer images were obtained from the ventral part of
the kidney (most proximal to the transducer) and assessed if available. 8 linear transducer
sets were not completely obtainable or insufficient for analysis due to non-compliance.
ln total, 40 curved transducer and 32 linear transducer datasets were analyzed.
Image analysis
The obtained datasets were analyzed in randomized order in consensus reading by two
radiologists (J. H., 12 years of renal transplant ultrasound experience; E. D. 1 year
of renal transplant ultrasound experience). Transplant vascularization was evaluated
in four categories:
Delineation of the entire renal vascular tree (Grade 1 – clear demarcation of interlobar,
together with arcuate and interlobular vessels; Grade 2 – clear demarcation of interlobar
and cortical vessels, but no distinction of interlobular from arcuate vessels; Grade
3 – only clear demarcation of interlobar vessels, Grade 4 – insufficient demarcation;
see [Fig. 1 ]);
Vessel density within the external cortex (interlobular vessels) in the ventral, dorsal,
and lateral aspect of the kidney (Grade 1 – high density with close vessel alignment;
Grade 2 – reduced vessel density with presence of small avascular gaps; Grade 3 –
low vessel density with dominant large avascular gaps or absence of vessels; see [Fig. 2 ]) [20 ]
[21 ].
Vessel-capsule distance (in cm); distance from the nearest visible cortical vessel
to the renal capsule. Measurement from cutis to renal capsule (a) on CDS image with
max. velocity range of 5 cm/s and from cutis to nearest vessel pixel on CDS (b) and
on corresponding BFS image (b); vessel-capsule distance equals ((b)–(a)) [22 ]
[23 ].
Cortical vessel count; number of distinguishable interlobular vessels in a length
of 1 cm on corresponding CDS and BFS images (see [Fig. 3 ]). For standardization, measurement was performed perpendicular to the point where
the vessel-capsule distance was assessed.
Linear transducer images were evaluated for categories 1–3, excluding the dorsal and
lateral renal cortex as not depicted on the image.
Fig. 1 7-year-old boy and corresponding ultrasound images of a kidney transplant in the
longitudinal plane performed with a curved transducer using the CDS (a–c , max. velocity ranges 15 cm/s, 9 cm/s and 5 cm/s) and the BFS technique d are shown. Note the absence of vessels in the lower kidney due to an occluded accessory
lower pole renal artery (c, d ; asterisk). The upper and middle parts of the kidney are well vascularized. Lowering
max. velocity ranges on CDS images a–c increases cortical vascular density but diminishes vessel delineation by aliasing
and blurring artifacts. BFS d better depicts the entire vascular tree (interlobar vessel arrow, arcuate vessel
dotted arrow, interlobular vessel arrowhead; renal vascular tree: grade 1) than CDS
(a–c ; renal vascular tree: grade 2). CDS has higher sensitivity for the detection of vessels
in the deeper structures (cortical vessel density in the ventral aspect of the kidney:
both grade 1; in the lateral aspect: CDS grade 1 and BFS grade 2; in the dorsal aspect:
both grade 3). Technical parameters: a. Pulse repetition frequency (PRF) 1.4, b. PRF
0.8, c. PRF 0.5, a.-c. Mechanical index (MI) 1.2, Thermal index in soft tissue (TIs)
0.9, Frequency (Frq) 3.6, Coded harmonic imaging (CHI) Frq 6.0, d. Pulse repetition
interval (PRI) 12, MI 1.2, TIs 1.7. CDS = color Doppler sonography; BFS = B-flow sonography.Abb. 1 7-jähriger Junge und zugehörige Ultraschallbilder eines Nierentransplantats in Longitudinalschnitt
mittels Sektorschallkopfes in CDS- (a–c , max. Geschwindigkeitsbereich 15 cm/s, 9 cm/s und 5 cm/s) und BFS-Technik d . Bemerke die Gefäßabwesenheit im kaudalen Nierenabschnitt aufgrund einer verschlossenen
akzessorischen Unterpolarterie (c, d ; Stern) Der kraniale und mittlere Nierenabschnitt zeigen sich regelrecht durchblutet.
Das Herabsetzen der max. Geschwindigkeitsbereiche auf den CDS-Bildern a–c steigert die kortikale Gefäßdichte, schränkt die Gefäßdifferenzierbarkeit jedoch
durch Aliasing und Unschärfeartefakte ein. BFS d stellt den gesamten Gefäßbaum besser dar (interlobäres Gefäß Pfeil, Bogengefäß gepunkteter
Pfeil, interlobuläres Gefäß Pfeilkopf; renaler Gefäßbaum: Grad 1) als CDS (a–c ; Grad 2). CDS hat eine höhere Sensitivität zur Gefäßdetektion in den tiefergelegenen
Nierenabschnitten (Dichte der Kortexgefäße am ventral gelegenen Nierenanteil: beide
Grad 1; lateral: CDS Grad 1, BFS Grad 2; dorsal: beide Grad 3). Technische Parameter:
a Pulsrepetitionsfrequenz (PRF) 1,4, b PRF 0,8, c PRF 0,5, a–c mechanischer Index
(MI) 1,2, thermischer Index in Weichteilen (TIs) 0,9, Frequenz (Frq) 3,6, Coded-harmonic-imaging
(CHI) -Frq 6,0, d Pulsrepetitionsintervall (PRI) 12, MI 1,2, TIs 1,7. CDS = Color-Doppler-Sonografie;
BFS = B-Flow-Sonografie
Fig. 2 10-year-old girl and corresponding ultrasound images with CDS (a , max. velocity range 5 cm/s) and BFS b of a kidney transplant positioned in the right flank. With CDS, large vascular-free
gaps were noted in the ventral cortex (arrowhead; cortical vessel density grade 3),
whereas the corresponding BFS image continuously depicts vessels in these areas (grade
1). Advantages of CDS can be seen in the dorsal and lateral part of the transplant
with better detectability of vessels (arrow; cortical vessel density grade 1 and 1
with CDS and 3 and 3 with BFS). CDS = color Doppler sonography; BFS = B-flow sonography.Abb. 2 10-jähriges Mädchen und zugehörige Ultraschallbilder mit CDS (a , max. Geschwindigkeitsgrenze 5 cm/s) und BFS b eines in der rechten Flanke positionierten Nierentransplantats. Mit CDS zeigen sich
große avaskuläre Lücken im ventralen Kortex (Pfeilkopf; kortikale Gefäßdichte Grad
3), während das korrespondierende BFS-Bild eine kontinuierliche Gefäßanreihung in
diesen Abschnitten darstellt (Grad 1). Vorteile von CDS erkennt man im dorsalen und
lateralen Transplantatabschnitt mit besserer Gefäßnachweisbarkeit (Pfeil; kortikale
Gefäßdichte Grad 1 bzw. 1 mit CDS und 3 bzw. 3 mit BFS). CDS = Color-Doppler-Sonografie;
BFS = B-Flow-Sonografie.
Fig. 3 Detail magnification of the renal cortex in a 7-year-old patient after kidney transplantation
(same patient as in Fig. 1) to illustrate vessel count performed in a region of interest
with a length of 1 cm (box). On CDS (left image) the interlobular vessels are masked
by aliasing and single vessels can only be identified by an opposing flow direction
(max. vessel count: 7). The higher spatial resolution of BFS allows clearer separation
of neighboring cortical vessels (max. vessel count: 12).Abb. 3 Detailvergrößerung des Nierenkortex bei einem 7-jährigen Patienten nach Nierentransplantation
(gleicher Patient aus Abb. 1). Auf einer Strecke von 1 cm (Kasten) erfolgte die Auszählung
differenzierbarer Kortexgefäße. Mit CDS (linkes Bild) werden einzelne Interlobulärgefäße
durch Aliasing maskiert und sind nur durch entgegengesetzte Flussrichtungen zu differenzieren
(max. 7 Gefäße). Die höhere räumliche Auflösung von BFS erlaubt eine deutlichere Separation
benachbarter Kortexgefäße (max. 12 Gefäße).
Statistical Analysis
The corresponding datasets for CDS and BFS were compared using Fisher’s exact tests
for categorical variables (categories 1 and 2) and paired sample t-tests for numeric
variables (categories 3 and 4). Normal distribution was tested for the numeric variables.
If not stated otherwise, categorical variables are provided as numbers and percentages,
continuous variables as mean ± standard deviation (SD). The effect of body mass index
(BMI) and patient age (years) on the imaging results was calculated by ordinal regression
(categories 1 and 2) and by a general linear model (categories 3 and 4). A P-value
< 0.05 was considered statistically significant. All statistical analyses were computed
with MedCalc for Windows (Mariakerke, Belgium), Excel (Microsoft Corporation, Redmond
WA, USA), and SPSS (Version24, IBM, Armonk, USA).
RESULTS
Curved transducer
Delineation of the entire renal vascular tree was superior with BFS compared with
the velocity-optimized CDS images (grade 1.3 ± 0.66 vs. 2.48 ± 0.6, p < 0.001) ([Fig. 1 ], [4 ]). Also, vessel density and differentiability in the ventral external renal cortex
was higher with BFS than with CDS (vessel density, grade 1.63 ± 0.59 vs. 2.05 ± 0.55,
p = 0.01; cortical vessel count, 8.05 ± 2.85 vessels vs. 5.65 ± 2.3 vessels, p < 0.001; [Fig. 2 ], [3 ], [4 ], [5 ]). The minimal vessel-capsule distance indicating the degree of vascularization of
the peripheral cortex was lower with BFS compared with CDS (0.16 cm ± 0.13 cm vs.
0.31 cm ± 0.15 cm, p < 0.001; [Fig. 5 ]). More distant from the transducer, in the dorsal and lateral aspect of the kidney
graft, BFS was less sensitive than CDS (cortical vessel density, grade 2.65 ± 0.53
vs. 1.80 ± 0.61 and 2.64 ± 0.48 vs 2 ± 0.64; each p < 0.001; [Fig. 2 ], [4 ]). All data regarding the curved transducer are summarized in [Table 2 ]. The delineation of the renal vascular tree with CDS was superior in patients with
a lower BMI (p = 0.04). No significant effect of BMI or age on all other CDS parameters
(vessel density, vessel-capsule distance, cortical vessel count) or on all BFS parameters
was found.
Fig. 4 Grading of the renal vascular tree and cortical vessel density within the ventral,
dorsal and lateral aspect of the kidney transplant on curved array images. Data are
mean values and standard deviations. CDS = color Doppler sonography; BFS = B-flow
sonography. Renal vascular tree: Grade 1 – clear demarcation of all vessels; Grade
2 – clear demarcation of interlobar from cortical vessels, no distinction of interlobular
from arcuate vessels; Grade 3 – just clear demarcation of interlobar vessels, Grade
4 – insufficient demarcation. Cortical vessel density: Grade 1 – high vessel density
with close alignment; Grade 2 – reduction of vessel density with avascular gaps; Grade
3 – large intervals without vascularity or absence of cortical vessels. *Statistically
significant.Abb. 4 Evaluation des renalen Gefäßbaums und der kortikalen Gefäßdichte im ventralen, dorsalen
und lateralen Nierentransplantatabschnitt auf Sektorschallkopfbildern. Die Daten sind
angegeben als Mittelwert und Standardabweichung. CDS = Color-Doppler-Sonografie; BFS = B-Flow-Sonografie.
Renal vascular tree (renaler Gefäßbaum): Grad 1 – eindeutige Abgrenzbarkeit der gesamten
Gefäße; Grad 2 – eindeutige Abgrenzbarkeit von Interlobär- und Kortexgefäßen, ohne
Differenzierbarkeit zwischen Interlobulär- und Arcuataegefäß; Grad 3 – lediglich gute
Abgrenzbarkeit der Interlobärgefäße; Grad 4 – insuffiziente Differenzierbarkeit. Cortical
vessel density (Dichte der Kortexgefäße): Grad 1 – hohe Gefäßdichte mit enger Aneinanderreihung;
Grad 2 – reduzierte Gefäßdichte mit avaskulären Lücken; Grad 3 – große avaskuläre
Intervalle oder Fehlen kortikaler Gefäße. *statistisch significant.
Fig. 5 Assessment of vessel-capsule distance and the best cortical vessel count comparing
BFS and CDS within the ventral cortex of the kidney transplant on curved array images.
Data are mean values and standard deviations. CDS = color Doppler sonography; BFS = B-flow
sonography. Vessel-capsule distance: distance from the most external vessel to the
renal capsule in cm. Cortical vessel count: Best cortical vessel count in the ventral
renal cortex with a length of 1 cm. *Statistically significant.Abb. 5 Evaluation des Gefäß-Kapsel-Abstands und der besten kortikalen Gefäßauszählung im
ventralen Nierenkortex auf Sektorschallkopfbildern verglichen zwischen BFS und CDS. Die
Daten sind angegeben als Mittelwert und Standardabweichung. CDS = Color-Doppler-Sonografie;
BFS = B-Flow-Sonografie. Vessel-capsule distance (Gefäß-Kapsel-Abstand): Abstand des
externsten kortikalen Gefäßes zur Nierenkapsel in cm. Cortical vessel count (kortikale
Gefäßzählung): maximale Anzahl kortikaler Gefäße auf 1 cm.
Table 2
Comparison of vessel delineation in kidney transplants with the BFS and the CDS technique
using a curved transducer.Tab. 2 Vergleich der Gefäßdarstellung in Nierentransplantaten mit BFS- und CDS-Technik unter
Verwendung eines Sektorschallkopfes.
parameter
BFS
CDS
p-value
renal vascular tree (grades 1–4)
1.33 ± 0.66
2.48 ± 0.6
< 0.001
cortical vessel density (grades 1–3)
1.63 ± 0.59
2.05 ± 0.55
0.01
2.65 ± 0.53
1.80 ± 0.61
< 0.001
2.64 ± 0.48
2 ± 0.64
< 0.001
vessel-capsule distance (cm)
0.16 ± 0.13
0.31 ± 0.15
< 0.001
cortical vessel count (no.)
8.05 ± 2.85
5.65 ± 2.30
< 0.001
Data are mean values and standard deviations. CDS = color Doppler sonography; BFS = B-flow
sonography. Renal vascular tree: Grade 1 – clear demarcation of all vessels; Grade
2 – clear demarcation of interlobar from cortical vessels, no distinction of interlobular
from arcuate vessels; Grade 3 – just clear demarcation of interlobar vessels, Grade
4 – insufficient differentiability. Cortical vessel density: Grade 1 – high vessel
density with close alignment; Grade 2 – reduction of vessel density with avascular
gaps; Grade 3 – large intervals without vascularity or absence of cortical vessels.
Vessel-capsule distance: distance from the most external vessel to the renal capsule
in cm. Cortical vessel count: Best cortical vessel count in the external renal cortex
with a length of 1 cm.
Linear transducer
The analysis of the linear transducer images demonstrated no significant differences
between CDS and BFS with respect to the delineation of the vascular tree, of the cortical
vessels or the vessel-capsule distance (see [Fig. 6 ], [7 ]
[Table 3 ]).
Fig. 6 7-year-old boy (same as in Fig. 1, 3) with kidney transplant positioned in the right
flank. Corresponding ultrasound images in the longitudinal plane with a linear transducer
using the B-mode a , the CDS b , and the BFS technique c are shown. Nearly complete absence of cortical vessels in the right lower part of
the cortex is shown due to occlusion of an accessory renal artery (arrowheads). The
upper left part of the cortex has good vascularization, equally shown in the CDS and
the BFS technique. CDS = color Doppler sonography; BFS = B-flow sonography.Abb. 6 7-jähriger Junge (derselbe wie in Abb. 1, 3) mit in der rechten Flanke positioniertem
Nierentransplantat. Korrespondierende Ultraschallbilder in Longitudinalschnitt unter
Verwendung eines Linearschallkopfes in B-Mode- a , CDS- b und BFS-Technik c . Nahezu vollständiges Fehlen von Kortexgefäßen im rechten, kaudalen Kortex aufgrund
einer verschlossenen akzessorischen Unterpolarterie (Pfeilköpfe). Der kraniale, linke
Abschnitt ist regelrecht perfundiert, ebenso gut mit CDS wie mit BFS dargestellt.
CDS = Color-Doppler-Sonografie; BFS = B-Flow-Sonografie.
Fig. 7 Assessment of vascularity comparing BFS and CDS when applying a linear transducer.
No statistical significance was found. Data are mean values and standard deviations.
CDS = color Doppler sonography; BFS = B-flow sonography. Renal vascular tree: Grade
1 – clear demarcation of all vessels; Grade 2 – clear demarcation of interlobar from
cortical vessels, no distinction of interlobular from arcuate vessels; Grade 3 – just
clear demarcation of interlobar vessels, Grade 4 – insufficient differentiability.
Cortical vessel density: Grade 1 – high vessel density with close alignment; Grade
2 – reduction of vessel density with avascular gaps; Grade 3 – large intervals without
vascularity or absence of cortical vessels. Vessel-capsule distance: distance from
the most external vessel to the renal capsule in cm.Abb. 7 Evaluation der renalen Gefäßversorgung mit BFS verglichen mit CDS bei der Verwendung
eines Linearschallkopfes. Wir fanden keine statistische Signifikanz. Die Daten sind
angegeben als Mittelwert und Standardabweichung. CDS = Color-Doppler-Sonografie; BFS = B-Flow-Sonografie.
Renal vascular tree (renaler Gefäßbaum): Grad 1 – eindeutige Abgrenzbarkeit der gesamten
Gefäße; Grad 2 – eindeutige Abgrenzbarkeit von Interlobär- und Kortexgefäßen, ohne
Differenzierbarkeit zwischen Interlobulär- und Arcuataegefäßen; Grad 3 – lediglich
gute Abgrenzbarkeit der Interlobärgefäße; Grad 4 – insuffiziente Differenzierbarkeit.
Cortical vessel density (Dichte der Kortexgefäße): Grad 1 – hohe Gefäßdichte mit enger
Aneinanderreihung; Grad 2 – reduzierte Gefäßdichte mit avaskulären Lücken; Grad 3
– große avaskuläre Intervalle oder Fehlen kortikaler Gefäße. Vessel-capsule distance
(Gefäß-Kapsel-Abstand): Abstand des externsten kortikalen Gefäßes zur Nierenkapsel
in cm. *statistisch signifikant
Table 3
Comparison of vessel delineation in kidney transplants with the BFS and the CDS technique
using a linear transducer.Tab. 3 Vergleich der Gefäßdarstellung in Nierentransplantaten mit BFS- und CDS-Technik unter
Verwendung eines Linearschallkopfes.
parameter
BFS
CDS
p-value
renal vascular tree (grades 1–4)
1.45 ± 0.68
1.26 ± 0.58
0.31
cortical vessel density in the ventral external cortex (grades 1–3)
1.77 ± 0.56
1.74 ± 0.52
0.92
vessel-capsule distance (cm)
0.09 ± 0.06
0.09 ± 0.05
0.61
Data are mean values and standard deviations. CDS = color Doppler sonography; BFS = B-flow
sonography. Renal vascular tree: Grade 1 – clear demarcation of all vessels; Grade
2 – clear demarcation of interlobar from cortical vessels, no distinction of interlobular
from arcuate vessels; Grade 3 – just clear demarcation of interlobar vessels, Grade
4 – insufficient differentiability. Cortical vessel density: Grade 1 – high vessel
density with close alignment; Grade 2 – reduction of vessel density with avascular
gaps; Grade 3 – large intervals without vascularity or absence of cortical vessels.
Vessel-capsule distance: distance from the most external vessel to the renal capsule
in cm.
Discussion
This study on kidney transplantation in children found that imaging of vascularization
can be substantially improved by adding BFS as a non-Doppler-based vascular imaging
technique to a standard protocol. The degree of transplant vascularization is a measure
of transplant viability and impairment follows acute and chronic functional disorders
[1 ]
[4 ]
[5 ].
As the quality of vessel delineation also depends on the applied ultrasound technique,
these methodical aspects have to be carefully controlled [20 ]
[23 ]
[24 ]
[25 ]
[26 ]. In a preliminary study with renal transplantation in adults, Russo et al. showed
that BFS compared with Doppler-based techniques can depict the cortical vasculature
more clearly and thus characterize causes of vascular complications more precisely
[16 ]. Further studies have to be performed in children to test clinical meaningfulness
and capability to monitor transplant viability.
The advantages of BFS can be attributed to the inherent properties of B-mode imaging.
Based on subtraction of B-mode images, flow images with high spatial and temporal
resolution can be generated [11 ]. In comparison with CDS, BFS was substantially better for the delineation of the
entire renal vascular tree allowing a more detailed depiction of the cortical vascular
architecture together with the feeding segmental arteries on a single preset. In our
clinic, we use the high spatial resolution of BFS for subsequent exact and quick placement
of spectral Doppler volumes. BFS shows a dynamic range to capture fast-flow and low-flow
vessels on a single flow image, as initially observed by Wachsberg et al. [12 ]. In contrast, realistic visualization of the renal vasculature is technically more
demanding and time-consuming with CDS, where velocity encoding sensitivities need
to be adapted to the region of interest with different settings for sensitive detection
of low-flow cortical vessels or fast-flow feeding arteries to avoid aliasing artifacts
[2 ]
[12 ]. We showed that BFS can separate the densely packed interlobular and arcuate vessels
of the external renal cortex, whereas in CDS blurring boundaries of neighboring vessels
were noted and explained by blooming artifacts.
On the other hand, BFS is prone to sound beam attenuation, which increases with depth
[10 ]. Applying B-flow on kidney grafts in children is favorable as the organ is situated
close to the skin in the iliac fossa with little adjacent subcutaneous fat [2 ]. Yet, also in our pediatric cohort, BFS was less sensitive than color Doppler for
vessel depiction in the deeper areas more distant from the center of the transducer.
Additionally, we did not notice significant differences between CDS and the BFS grading
of transplant vascularity when using a linear, high-resolution transducer. As CDS
in overlay mode also carries information about structural parenchymal integrity, e. g.
can detect cortical scarring, we favor CDS for this application. However, due to their
lower display range, linear transducers are limited to assessment of accessible regions
of the transplant, e. g. the superficial part of the lower pole.
There are further methods recently introduced by different manufacturers that seem
advantageous for the sonographic depiction of complex flow and small vessels, such
as Advanced Dynamic Flow (ADF) or superb micro-vascular imaging (SMI) [27 ]
[28 ]. However, a common problem of these newer techniques, as well as of BFS, is the
lacking overall availability on the ultrasound systems. Also, contrast-enhanced ultrasound
(CEUS) is an increasingly used method for vascular assessment in kidney transplantation
in adults and has been recommended for this use by the European federation of Societies
in Ultrasound in Medicine and Biology (EFSUMB) [24 ]
[30 ]
[31 ]. To our knowledge, CEUS has not yet been systematically applied in pediatric kidney
transplantation as its intravascular application in children is only approved for
the characterization of focal liver lesions so far [29 ].
Our study has the following limitations: (1) The study design is retrospective and
is therefore dependent on medical documentation and principally prone to selection
bias. (2) Results of ultrasound investigations are operator-dependent and require
a high level of skill. To guarantee a high level of data consistency, the included
data points were limited to examinations performed by a single expert pediatric radiologist.
The inter- and intra-operator variability of BFS and CDS cannot be reported (3). The
number of patients in our study was relatively low and heterogeneous with regard to
age, days after transplantation, BMI, transplantation type (living or cadaveric),
clinical and laboratory findings ([Table 1 ]).
In summary, BFS yields better results than CDS for assessing the overall vascularity
in pediatric kidney transplantation and thereby could improve monitoring of transplant
viability. As suggested by previous studies focusing on other vascular territories,
the B-flow technique may be especially useful in infants and young children [18 ]
[19 ]. BFS is less favorable in larger patients and for the deeper parts of the kidney
due to sound beam attenuation and should thereby be used in addition to Doppler-based
ultrasound techniques. Further standardization of the ultrasound protocols and the
reporting of the results in pediatric kidney transplantation is desirable for the
future. There are other fields of diagnostic ultrasound, e. g. in the domain of screening
and surveillance where standardization is already more advanced [32 ].
Clinical Relevance of the study
Additional monitoring with BFS improves monitoring of kidney transplant viability.
Acute and chronic functional disorders impair transplant vascularity. A higher quality
of vessel delineation may facilitate the early detection of graft damage.
Adding BFS to a standard protocol after kidney transplantation accelerates workflow
as an accurate overview of the global vascular tree is obtained and subsequent detailed
vascular assessment can be performed, e. g. placement of spectral doppler volumes.
