Original Investigation

Assessment of Splenic Perfusion inPatients with Malignant Hematologic

Diseases and Spleen Involvement,Liver Cirrhosis and Controls Using

Volume Perfusion CT (VPCT):

A Pilot Study

Alexander W. Sauter, MD, Stefan Feldmann, MDent, Daniel Spira, MD, Maximilian Schulze, MD,Ernst Klotz, PhD, Wichard Vogel, MD, Claus D. Claussen, MD, Marius S. Horger, MD

Ac

FrD.HeSeFo13sp

ªdo

Rationale and Objectives: The aim of this study was to assess splenic perfusion in patients with spleen involvement in malignanthematologic diseases and liver cirrhosis and in controls without hepatosplenic disease using volume perfusion computed tomography.

Materials and Methods: Between October 2009 and December 2011, 14 hematologic patients with known spleen involvement were

recruited. An additional 17 consecutive patients without known splenic or liver disease were enrolled as controls, as well as 29 patientswith liver cirrhosis and portal hypertension. A 40-second volume perfusion computed tomographic scan of the upper abdomen was

performed. Analysis included measurement of splenic volume, blood flow (BF), blood volume (BV), Ktrans, and mean transit time (MTT).

Results: In lymphoma patients, mean splenic volume and perfusion parameters were as follows: splenic volume, 1125.34 mL; BF,61.24 mL/100 mL/min; BV, 16.53 mL/100 mL; Ktrans, 37.00 mL/100 mL/min; and MTT, 12.42 seconds. All perfusion values of patients

with lymphoma and cirrhosis differed significantly, except for BV, compared to controls. For patients with lymphoma, significant correla-

tions were found between splenic volume and BF (r =�0.683, P = .000), splenic volume and BV (r =�0.525, P = .002), and splenic volume

andMTT (r = 0.543,P = .001). During treatment, significant correlations between the diameters of nodular lymphoma target lesions, splenicvolume, and the perfusion parameters were present for splenic volume (r = 0.601, P = .002), BF (r = �0.777, P = .000) and BV (r = �0.500,

P = .011).

Conclusions: Volume perfusion computed tomography represents a novel tool for the assessment of splenic perfusion. Preliminaryresults in patients with spleen involvement reveal lower perfusion values compared to controls or patients with cirrhosis. Therefore, this

technique might provide additional information in clinical routine.

Key Words: CT perfusion; volume-perfusion computed tomography; spleen; lymphoma; liver cirrhosis.

ªAUR, 2012

As a result of rare primary splenic diseases (1), the spleen

is often considered ‘‘silent and forgotten’’ (2). On the

other hand, secondary manifestations of hematologic,

immunologic, oncologic, infectious, vascular, and systemic

disorders are frequently found in the spleen (2). Forty-four

ad Radiol 2012; -:1–9

om the Department of Diagnostic and Interventional Radiology (A.W.S., S.F.,S., M.S., C.D.C., M.S.H.) and the Medical Center, Department ofmatology and Oncology (W.V.), University Hospital of T€ubingen, Hoppe-yler-Straße 3, D-72076 T€ubingen, Germany; and Siemens Healthcare,rchheim, Germany (E.K.). Received September 2, 2011; accepted January, 2012. This study was supported by Siemens Healthcare. Address corre-ondence to: A.W.S. e-mail: alexander.sauter@klinikum.uni-tuebingen.de

AUR, 2012i:10.1016/j.acra.2012.01.007

percent of patients with lymphoma have splenic involvement

(3). Splenomegaly is a common but unspecific finding, as

enlarged spleens do not necessarily indicate lymphoma infil-

tration (4), and only 1% to 2% of all patients with non-

Hodgkin’s lymphoma present with such splenic enlargement

(5). Spleen involvement is common in mantle cell lymphoma

and follicular lymphoma (5,6), as well as chronic lymphatic

leukemia (7–9). Accurate staging with assessment of splenic

disease is essential for clinical management and follow-up

(10). However, a study by Munker et al (11) revealed a low

accuracy of only 15% to 37% for computed tomography

(CT) and a disappointing correlation with histologic state.

Nevertheless, CT is the most frequently used stagingmodality.

Especially for diffuse involvements, positron emission

1

SAUTER ET AL Academic Radiology, Vol -, No -, - 2012

tomography (PET) using [18F]-fluorodeoxyglucose (FDG)

seems to be superior to CT (12). Secondary portal hyperten-

sion due to either direct lymphoma infiltration of the liver (13)

or mediastinal compression of the superior vena cava (14) may

additionally alter splenic perfusion. In a similar scenario, portal

hypertension in liver cirrhosis can prolong blood transit time

because of splenic venous congestion and decreased splenic

perfusion (15). Various imaging techniques such as isotope

scintigraphy (16–18) and [15O]-labeled water PET (19) have

been applied for the assessment of hepatic and splenic perfu-

sion. These techniques are complex and are available only

at specialized imaging sites. Functional CT, based on the

exchange of iodinated contrast material between the intravas-

cular space and the extravascular interstitial space, is an inter-

esting approach that can be integrated in the clinical routine

with manageable effort (20). This method evolved from

single-location dynamic sequences (15,21,22) toward multi–

detector row computed tomographic perfusion (23). Recently,

Goetti et al (24) applied four-dimensional spiral-mode for

computed tomographic liver perfusion with 128-row CT and

coverage of up to 14.8 cm.

The aim of our study was to assess and compare the ranges

of splenic perfusion using this new method. In the first step,

a group of hematologic patients with spleen involvement

was analyzed. Elderly patients in particular require special

attention, because in this group, liver cirrhosis is frequently

(23.7%) undiagnosed (25). Therefore, we additionally

analyzed splenic perfusion in a population with proven liver

cirrhosis, portal hypertension and hepatocellular carcinoma

(HCC) who had undergone liver perfusion CT prior to trans-

arterial chemoembolization. Finally, a group of patients who

had undergone tumor perfusion CT for other purposes but

presented with neither splenic diseases nor portal hyperten-

sion were included as normal controls.

To our knowledge, this is the first report in the literature

dealing with the role of volume perfusion CT (VPCT) for

the assessment of splenic perfusion differences in patients

with lymphomatous spleen involvement, patients with

cirrhosis, and controls.

MATERIALS AND METHODS

Study Population

Patients aged $30 years were enrolled consecutively from

October 2009 through December 2011 and analyzed retro-

spectively. Data were prospectively acquired in previous

perfusion studies approved by the local ethics committee,

including therapy monitoring in patients with lymphoma,

HCC, and general cancer disease with palliative care. Exclu-

sion criteria for contrast application included kidney dysfunc-

tion (defined as a serum creatinine level > 150 mmol/L),

known hypersensitivity to iodine-containing contrast media,

pregnancy, and untreated hyperthyroidism. All patients

provided written informed consent, including information

about radiation exposure.

2

For the first group, 14 patients (four women, 10 men; mean

age, 61.7 � 8.3 years; range, 49.5–76.8 years) with hemato-

logic diseases (B-cell chronic lymphocytic leukemia, n = 3;

B-cell non-Hodgkin lymphoma, n = 3; diffuse large B-cell

lymphoma, n = 2; peripheral T-cell lymphoma not otherwise

specified, n = 3; mantle cell lymphoma, n = 2; splenic

marginal zone lymphoma, n = 1) with spleen involvement

were examined. These patients were assessed either for initial

diagnosis or for detection of progressive disease. Follow-up

perfusion measurements were performed in 9 patients.

Evaluation of the disease course was accomplished using

Cheson criteria considering all measurable lymphoma

manifestations.

The second group included 29 patients (four women,

25 men; mean age, 65.0 � 8.6 years; range, 46.0–78.0 years)

with liver cirrhosis and HCC. Twenty-two patients were

classified as Child-Turcotte-Pugh class A and seven patients

as class B. All patients had signs of liver cirrhosis and portal

hypertension on contrast-enhanced CT (irregular and

nodular surface, blunt edge, parenchymal abnormalities,

morphologic changes [atrophy of the right lobe, atrophy of

the medial segment, hypertrophy of the lateral segment,

hypertrophy of the caudate lobe, widened pericholecystic

space, enlarged periportal space], manifestations of portal

hypertension [splenomegaly, splenic vein dilatation, ascites,

gastric and esophageal varices, paraumbilical collaterals, sple-

norenal shunt, other sites of collaterals] [26]) and had under-

gone liver perfusion CT for evaluation prior to transarterial

chemoembolization. Patients with portal vein thrombosis

and transjugular intrahepatic portosystemic shunts were

excluded.

Finally, 17 patients (five women, 12 men; mean age, 59.7�14.4 years; range, 34.6–78.1 years) who had been examined

with perfusion CT because of tumor manifestation in the

upper abdomen without hepatic or splenic spread were

included. These patients had the following diagnoses:

non-small-cell lung cancer (n = 1), pancreatic cancer

(n = 4), cancer of unknown primary (n = 1), esophageal

cancer (n = 4), renal cell carcinoma (n = 1), leiomyosarcoma

(n = 2), gastric cancer (n = 1), colorectal cancer (n = 1),

nondysgerminoma (n = 1), and Echinococcus granulosus

(n = 1). In this group, patients with signs of liver cirrhosis

on contrast-enhanced CTwere excluded.

Computed Tomographic Protocol

All examinations were performed with a 128-row computed

tomographic scanner (Somatom Definition AS+; Siemens

Healthcare, Forchheim, Germany). First, non-enhanced

low-dose CT of the thorax and abdomen was performed.

Subsequently, an experienced radiologist analyzed the non-

enhanced low-dose computed tomographic findings, and

a scan range of 6.9 to 9.6 cm z-axis coverage centered to the

spleen hilum was planned, followed by VPCTusing an adap-

tive spiral scanning technique. The following scan parameters

were used: tube voltage, 80 kV; collimation, 128 � 0.6 mm;

Academic Radiology, Vol -, No -, - 2012 SPLENIC PERFUSION

and a total of 22 scans. Depending on the bodyweight, patients

were examinedwith tube current–time products of 80, 100, or

120 mAs. The total scanning time for VPCTwas 40 seconds.

During perfusion scanning, patients were asked to resume

shallow breathing for the entire duration of the study. A

volume of 50 mL Ultravist 370 (Bayer Vital, Leverkusen,

Germany) at a flow rate of 5 mL/s was injected in an

antecubital vein through an 18-gauge needle (Vasofix; B.

Braun Melsungen AG, Melsungen, Germany) followed by

a saline flush of 50 mL sodium chloride at 5 mL/s. Scanning

was started after a delay of 7 seconds. Contrast material was

administered using a dual-head pump injector (Medtron, Saar-

bruecken,Germany). After the perfusion scans, another 80mL

of contrast medium was injected, and chest, abdominal, and

neck computed tomographic scans for staging purposes were

subsequently performed. One set of axial images with a slice

thickness of 3 mm for perfusion analysis was reconstructed

without overlap, using a medium-smooth tissue convolution

kernel (B10f). All images were then anonymized and trans-

ferred to an external workstation (Multi-ModalityWorkplace;

Siemens Healthcare) for analysis.

Quantitative Perfusion Analysis

Data evaluation was performed using syngo Volume Perfusion

CT Body (Siemens Healthcare), on the basis of the Tofts

model. No clinical data except gender and age were provided

to the reader. Motion correction and noise reduction were

performed for all image sets using the algorithms integrated

in the perfusion postprocessing software. The algorithms

are based on non-rigid deformable registration for anatomic

alignment (27,28) and a dedicated noise reduction

technique for dynamic data (29).

A volume of interest was drawn manually around the entire

spleen on the maximum intensity projection image set. Care

was taken to exclude hilar blood vessels. Navigation

throughout all slices was performed to ensure a good selection

result. Air, fat, bone, and other dense structures were excluded

with a threshold-based segmentation. For the arterial input

function, a region of interest was placed inside the abdominal

aorta. Perfusion parameter volume maps of blood flow (BF),

blood volume (BV), Ktrans, and mean transit time (MTT)

were generated. Splenic volume was determined using syngo

Oncology (Siemens Healthcare) from portal venous–phase

enhanced abdominal scans.

Statistical Analyses

Descriptive statistics including means and standard deviations

of the splenic volumes and perfusion parameters were calcu-

lated using JMP version 8.0.2 (SAS Institute Inc, Cary,

NC). Differences between the perfusion parameters of the

groups were tested using unpaired t tests. Correlations were

analyzed using Spearman’s rank correlation coefficient (r).

Differences were considered to be significant for two-tailed

P values <.05.

RESULTS

All perfusion studies were completed successfully and could be

analyzed using the software.

The mean dose-length products for perfusion measurements

in the upper abdomen were 480.0 � 97.5 mGy $ cm for the

spleen involvement group, 398.0 � 234.5 mGy $ cm for the

cirrhotic group, and 369.4 � 93.6 mGy $ cm for the control

group.

Spleen Involvement

Results of perfusion measurements in patients with spleen

involvement by malignant hematologic diseases are summa-

rized in Table 1. Table 2 provides a detailed parameter

assembly for each patient. Splenic involvement could be

confirmed in two patients with FDG-PET (patients 2 and

10) and in one patient with histology after splenectomy

(patient 2).

Mean splenic volume and perfusions parameters (BF,

BV, Ktrans, and MTT) were as follows: splenic volume,

1125.34 � 784.24 mL; BF, 61.24 � 22.80 mL/100 mL/min;

BV, 16.53 � 14.32 mL/100 mL; Ktrans, 37.00 � 12.97 mL/

100 mL/min; and MTT, 12.42 � 2.66 seconds. Splenic

volume (P < .001), BF (P < .001), Ktrans (P = .011), and

MTT (P = .023) differed significantly from the control group,

whereas BV (P = .095) did not. Compared to cirrhotic

patients, only Ktrans (P < .001) revealed a significant differ-

ence, whereas splenic volume (P = .077), BF (P = .114),

BV (P = .618), and MTT (P = .071) did not. Significant

correlations were found between splenic volume and BF

(r = �0.683, P = .000), splenic volume and BV

(r = �0.525, P = .002), and splenic volume and MTT

(r = 0.543, P = .001) but not for splenic volume and Ktrans

(r = 0.071, P = .698). Areas of focal spleen involvement

were present in only six patients on the basis of hypodense

areas in the portal venous–phase enhanced abdominal scans.

In patient 8, the hypodense lesions disappeared under therapy.

Nine patients were followed up during ongoing treatment

(see Table 2). As expected, follow-up perfusion studies

revealed decreases in BF and BV in patients experiencing

progressive disease (patient 8, May to August 2010; and

patient 11), whereas in seven patients (patient 7; patient 8,

November 2010 to March 2011; and patients 9, 10, 12, 13,

and 14) undergoing treatment, increases of BF and BV

associated with a decrease in splenic volume were registered.

Additionally, the diameters of nodular lymphoma target

lesions were correlated with splenic volume and the perfusion

parameters of the spleen. Interestingly, significant correlation

were present for splenic volume (r = 0.601, P = .002), BF

(r = �0.777, P = .000), and BV (r = �0.500, P = .011)

but not for Ktrans (r = �0.286, P = .166) or MTT

(r = 0.341, P = .096).

Figure 1a shows the BF map of patient 2, with peripheral

T-cell lymphoma not otherwise specified. The black stars

indicate an area of low BF. This area corresponds to increased

3

TABLE 1. Splenic Volumes and Perfusion Parameters for All Groups

Variable

Malignant

Hematologic

Diseases P*

Control

Group Py

Entire

Cirrhotic

HCC Group

Cirrhotic

HCC Child

Class A Group

Cirrhotic

HCC Child

Class B Group

Splenic volume (mL) 1125.34 � 784.24 <.001 253.79 � 75.28 <.001 639.09 � 355.22 594.56 � 375.96 779.05 � 253.04

Blood flow (mL/100 mL/min) 61.24 � 22.80 <.001 95.66 � 25.25 .002 72.38 � 21.00 74.51 � 22.93 65.67 � 12.24

Blood volume (mL/100 mL) 16.53 � 14.32 .095 16.39 � 4.64 .089 13.91 � 6.46 13.39 � 6.58 15.54 � 6.27

Ktrans (mL/100 mL/min) 37.00 � 12.97 .011 49.12 � 15.58 .036 52.21 � 10.43 51.84 � 11.19 53.38 � 8.22

Mean transit time (seconds) 12.42 � 2.66 .023 10.67 � 2.05 <.001 13.77 � 3.57 13.56 � 3.91 14.46 � 2.32

HCC, hepatocellular carcinoma.

Data are expressed as mean � standard deviation.

*Patients with malignant hematologic diseases versus controls.yControls versus entire cirrhotic HCC group.

SAUTER ET AL Academic Radiology, Vol -, No -, - 2012

FDG uptake (Fig 1b). Figures 1c to 1e present the maps for

BV, Ktrans, and MTT. No clear and circumscribed areas are

seen in these maps. Fast and slow enhancing compartments

of the spleen cause the heterogeneous patterns. The patient

underwent splenectomy, and the enlarged spleen was histo-

logically analyzed. Figure 1f shows a hematoxylin and eosin

section of the spleen. A diffuse infiltration was present, but

also fine nodular infiltrates (surrounded by black stars) could

be seen. The disease was so advanced that even some trabec-

ular arteries were infiltrated by lymphoma cells (Fig 1g).

Patients with Nonhepatosplenic Disease and LiverCirrhosis

The splenic volumes and perfusion parameters for the

cirrhotic HCC groups and control groups are shown in

Table 1.

Not surprisingly, the mean splenic volume of the cirrhotic

HCC group (639.09 � 355.22 mL) was higher than that of

the control group (253.79 � 75.28 mL) (P < .001). The

mean BF in the cirrhotic HCC group was 72.38 mL/100

mL/min, compared to 95.66 mL/100 mL/min in control

patients (P = .002). The mean BVs of the two groups

(13.91 and 16.39 mL/100 mL) did not differ significantly

(P = .089), while the mean Ktrans values (52.21 and 49.12

mL/100 mL/min) were different (P = .036). Also, the

MTTs showed a significant difference (13.77 vs 10.67

seconds, P < .001). No significant correlations were found

between splenic volumes and the perfusion parameters in

the control patients (splenic volume and BF: r = 0.123,

P = .632; splenic volume and BV: r = �0.037, P = .883;

splenic volume and Ktrans: r = �0.002, P = .989; and splenic

volume and MTT: r = �0.052, P = .839) and the cirrhosis

patients (splenic volume and BF: r=�0.151, P= .432; splenic

volume and BV: r = 0.238, P = .211; splenic volume and

Ktrans: r = 0.080, P = .676; and splenic volume and MTT:

r = 0.174, P = .362).

The cirrhotic HCC group can be further subdivided into

patients in Child classes A and B (see Table 1). Splenic volumes

(Child class A, 594.56 � 375.96 mL; Child class B, 779.05 �253.04 mL; P= .088), BF values (Child class A, 74.51� 22.93

4

mL/100 mL/min; Child class B, 65.67� 12.24 mL/100 mL/

min; P = .341), BV values (Child class A, 13.39 � 6.58 mL/

100 mL; Child class B, 15.54 � 6.27 mL/100 mL; P = .372),

Ktrans values (Child class A, 51.84 � 11.19 mL/100 mL/min;

Child class B, 53.38 � 8.22 mL/100 mL/min; P = .899), and

MTTs (Child class A, 13.56 � 3.91 seconds; Child class B,

14.46 � 2.32 seconds; P = .252) did not differ significantly.

DISCUSSION

In this study, we aimed to determine the potential benefit of

VPCT for the assessment of normal and pathologic splenic

perfusion, including detection of abnormalities related to

cell infiltration in malignant hematologic diseases.

The spleen is a blood-filled organ playing important roles

with regard to red blood cells and the immune system. It

represents the largest filter of blood and is organized as

a tree of branching arterial vessels with the arterioles ending

in a venous sinusoidal system (30). Splenic architecture

comprises many communicating compartments that build

up two distinct tissues, the red and the white pulp. Red

pulp is composed of nonanastomosing arterial vessels, thin-

walled venous vessels called splenic sinuses, splenic cords lying

between the sinusoids, and draining red pulp veins; white pulp

consists of lymphatic tissue (31). The dynamics of splenic

microcirculatory BF and the role of the spleen with respect

to red blood cells have been elucidated by means of several

different experimental approaches. Different microscopic

approaches have provided complementary information and

have clarified a number of important issues. Hence, exploiting

the quantitative analysis of BF and other perfusion character-

istics for spleen characterization using available and relatively

low-cost technology has the potential to provide new insights

into this complex and neglected organ (32).

A major focus in daily hemato-oncologic diagnosis is the

accurate detection of splenic involvement with malignant

hematologic diseases. Noninvasive assessment of splenic

involvement in lymphoma is important for both staging and

therapeutic purposes. In Hodgkin’s disease, for instance, the

spleen is the abdominal location most commonly affected by

occult disease and in 10% of the cases is the only site of

TABLE 2. Patients with Malignant Hematologic Diseases and Spleen Involvement

Patient Age (y) Sex Diagnosis First Diagnosis Treatment Study Date Splenic Volume BF BV Ktrans MTT

Baseline

1 58.6 M Low-grade B-NHL 02.2007 No 10.2010 1826.31 59.35 6.51 52.46 14.29

2 72.7 M Relapse PTCL-NOS 03.2011 6 � CHOP (5/06 to 8/06) 01.2011 560.21 87.69 13.97 47.60 10.54

3 68.6 F SMZL 12.2010 No 03.2010 1878.34 55.82 7.45 43.36 15.13

4 65.7 M B-CLL 07.2005 15 � KNOSPE (5/07 to 12/07)

1 � rituximab (6/08)

5 � R-bendamustine (6/08 to 10/08)

5 � fludarabine/cyclophosphamide

(9/09 to 1/10)

04.2011 648.92 46.76 8.37 30.78 12.16

5 76.8 M B-CLL 01.1998 5 � KNOSPE (12/01 to 3/02)

6 � KNOSPE (7/03 to 10/03)

4 � R-bendamustine (10/09 to 2/10)

06.2010 2057.63 51.62 14.09 55.17 16.86

Baseline and follow-up

6 49.5 M B-CLL 10.2005 No 01.2011 503.45 64.88 13.20 22.09 10.11

6 � R-FC (3/11 to 8/11) 09.2011 320.55 117.95 55.67 202.42 11.97

7 63.6 F Relapse low-grade

B-NHL

08.2010 6 � VACOP-B (2/95)

6 � R-bendamustine (8/07 to 2/08) 01.2011

922.71 25.40 3.56 28.36 13.66

2 � R-VIPE (1/11) 02.2011 457.40 73.98 15.42 40.21 11.29

8 56.6 M Low-grade B-NHL 10.2008 6 � R-CHOP + 2 � rituximab

(10/08 to 3/09)

05.2010 1187.25 45.53 17.43 24.15 19.00

1 � R-VIPE (10/10) 08.2010 1748.80 37.51 4.93 36.07 16.01

2 � R-VIPE (11/10) 11.2010 1168.38 43.23 6.83 39.53 13.28

No 12.2010 1168.83 45.08 6.53 41.28 12.44

4 � R-bendamustine (11/10 to 6/11) 03.2011 1116.84 37.83 5.50 29.26 12.76

9 56.6 F DLBCL 07.2010 No 07.2010 1886.14 13.00 3.13 12.46 8.79

4 � R-CHOP (7/10 to 9/10) 09.2010 448.37 60.91 9.72 37.55 11.61

2 � R-CHOP + 2 � rituximab

(10/10 to 12/10)

01.2011 329.28 65.88 9.89 47.66 9.72

No 06.2011 320.18 94.16 41.47 33.84 10.90

10 67.7 M Relapse DLBCL 12.2010 6 � CHOP-21 (7/02 to 11/02) 02.2011 513.55 48.09 7.62 35.62 13.64

1 � R-VIPE (2/11) 03.2011 275.24 99.54 17.33 37.40 8.30

2 � R-VIPE (3/11) 04.2011 279.67 79.95 14.80 36.09 7.71

1 � HD + PBSCT (4/11) 08.2011 260.00 69.21 10.50 37.63 9.16

11 51.5 M PTCL-NOS 05.2010 2 � CHOP + 2 � VIPE (5/10 to 8/10) 08.2010 1856.29 49.86 15.43 47.25 14.30

HD + PBSCT (10/10) 12.2010 2982.11 32.37 4.20 36.13 12.12

12 50.5 F PTCL-NOS 02.2003 No 04.2010 1862.25 50.93 6.66 46.75 14.98

3 � CHOP + 2 � VIPE + 2 � DHAP

+ HD-BEAM (4/10 to 2/11)

03.2011 1241.66 60.56 32.04 29.77 12.26

1 � cladribine (5/11) 06.2011 1017.42 83.80 48.92 4.73 15.15

(continued on next page)

Academic

Radiology,Vol-

,No-

,-

2012

SPLENIC

PERFUSIO

N5

TABLE2.(continued)

Patient

Age(y)

Sex

Diagnosis

FirstDiagnosis

Treatm

ent

StudyDate

Splenic

Volume

BF

BV

Ktrans

MTT

13

56.6

MMCL

11.2010

2�

R-C

HOP;1�

R-D

HAP

12.2010

560.85

85.30

39.98

25.89

11.98

1�

R-D

HAP

01.2011

407.80

95.46

50.64

14.11

12.76

1�

HD+PBSCT(4/11)

05.2011

323.11

86.82

18.76

44.48

10.84

14

60.6

MMCL

03.2010

No

03.2010

2908.62

48.95

6.18

40.93

10.26

3�

R-C

HOP+2�

R-D

HAP

(4/10to

7/10)

06.2010

1069.75

88.09

39.88

67.18

10.34

B-C

LL,B

-cellchroniclymphocyticleukemia;B

F,b

loodflow;B

-NHL,B

-cellnon-H

odgkin’slymphoma;B

V,b

loodvolume;D

LBCL,d

iffuse

largeB-celllymphoma;F

L,follicularlymphoma;H

D,

high-dosechemotherapy;KNOSPE,chlorambucil,prednisolone;MCL,mantlecelllymphoma;MTT,meantransittime;PBSCT,peripheralb

loodstem

celltransplantation;PTCL-N

OS,periph-

eralT-celllymphomanototherw

isespecified;R-C

HOP,rituxim

ab,cyclophosphamide,hydroxydaunorubicin,vincristine,prednisone;R-D

HAP,rituxim

ab,dexamethaso

ne,cytarabine,

cisplatin;R-FC,rituxim

ab,fludarabine,cyclophosphamide;R-VIPE,rituxim

ab,mesna,cisplatin,etoposide;SMZL,splenicmarginalzonelymphoma;T-N

HL,T-cellnon-H

odgkin’s

lymphoma;

VACOP-B

,etoposide,doxorubicin,cyclophosphamide,vincristine,prednisone,bleomycin.

SAUTER ET AL Academic Radiology, Vol -, No -, - 2012

6

infradiaphragmatic involvement (4). Consequently, splenic

involvement in Hodgkin’s disease is responsible for a higher

stage of disease and alteration in treatment strategies. Deter-

mination of disease extent is also critical for appropriate treat-

ment planning and determining prognosis in non-Hodgkin’s

lymphomas (12). The following pitfalls often complicate diag-

nosis with CT imaging: (1) diffuse infiltration can appear

homogenously, or lymphoma lesions can be beyond the

scan resolution and might cause a uniform enhancement;

(2) diffuse and focal lymphoma infiltration can be present

next to each other, and only the focal lesions might be

detected; and (3) infarction or arciform splenic enhancement

patterns can hamper differentiation from lymphoma

infiltration.

Hence, the aim of this study was to provide an up-to-date

overview of the diagnostic performance of CT, using

functional VPCT, in detecting splenic involvement by malig-

nant hematologic diseases compared to controls and patients

with liver cirrhosis.

Most computed tomographic studies were performed in

the era of single-detector scanners and 10-mm axial slices.

Using splenic index measurements, Strijk et al (33) achieved

100% accuracy for discrimination between involved and

noninvolved spleens. Other size-based computed tomo-

graphic parameters for the detection of splenic involvement

were proposed by Daskalogiannaki et al (34), including

splenic volume and percentage change in splenic volume

at follow-up, all of them increasing the sensitivity and spec-

ificity of this technique. Nevertheless, particularly in

diffusely infiltrated spleens, tumor detection remains chal-

lenging, because a huge variety of other pathologies,

including inflammation and immunologic reaction, can

mimic involvement.

As an alternative, FDG PET has gained wide acceptance as

a very useful imaging modality in evaluating a variety of

neoplastic and inflammatory or infectious diseases. The major

advantage of FDG PETover traditional anatomic imaging is

that it depicts metabolic abnormalities, which often precede

changes at the anatomic level (35). However, for both CT

and FDG PET, the detection of focal uptake in the spleen is

always easier compared to the identification of diffuse infiltra-

tion. The degree of FDG uptake ultimately depends on the

glucose metabolism, which is known to be reduced in low-

grade lymphoma (36). Thus, the problem of detecting spleen

involvement in lymphoma has not been completely solved,

especially for low-grade hematologic malignancies.

Because VPCT-based splenic perfusion measurements

involve a completely different principle compared to glucose

use by tumor cells, the hypothesis was that tumor cell infiltra-

tion of the spleen could influence BF kinetics. Additionally,

lymphoma can induce secondary portal hypertension by

mass compression of the superior or inferior vena cava (14)

or by hepatic involvement (13). To exclude this possibility,

we determined splenic perfusion parameters both in noncir-

rhotic and cirrhotic patients. Our results yielded significant

differences in BF (95.66 vs 72.38 mL/100 mL/min), Ktrans

Figure 1. (a) Blood flow map of a patient with

peripheral T-cell lymphoma not otherwise

specified (PTCL-NOS) with splenic involvement.The black stars indicate an area with low blood

flow. (b) Corresponding [18F]-fluorodeoxyglu-

cose positron emission tomographic image ofthe patient with PTCL-NOS showing increased

FDG uptake. (c)Blood volumemap of the patient

with PTCL-NOS. Fast and slow enhancing

compartments of the spleen cause the heteroge-neous patterns. (d) Ktrans map of the patient with

PTCL-NOS. Fast and slow enhancing compart-

ments of the spleen cause the heterogeneous

patterns. (e)Mean transit time map of the patientwith PTCL-NOS. Fast and slow enhancing

compartments of the spleen cause the heteroge-

neous patterns. (f) Hematoxylin and eosin (H&E)

staining of a spleen specimen from a patientwith PTCL-NOS (magnification, 5�). The black

stars indicate a fine nodular infiltrate and

surrounding diffuse infiltrations. (g) H&E stainingof a spleen specimen from a patient with PTCL-

NOS (magnification, 5�). The black star shows

lymphomatous invasion into a trabecular artery.

Academic Radiology, Vol -, No -, - 2012 SPLENIC PERFUSION

(52.21 vs 49.12 mL/100 mL/min), and MTT (10.67 vs 13.77

seconds) between controls and patients with cirrhosis. Blom-

ley et al (22) reported similar differences between patients with

cirrhosis and normal controls using a single-slice technique.

Usually, arciform splenic enhancement detected by standard

CT represents no diagnostic challenge. However, differentia-

tion can become difficult in cases with diffuse, uniform splenic

involvement. Consequently, we analyzed splenic perfusion

7

SAUTER ET AL Academic Radiology, Vol -, No -, - 2012

parameters in patients with known lymphomatous involve-

ment of the spleen to determine its impact on perfusion

parameters. Correspondingly, BF and BV were comparatively

reduced at initial presentation as well as in untreated relapsed

lymphoma. This phenomenon is presumed to be the conse-

quence of progressive sinusoidal compression by increasing

lymphoma cell population in the cords of the white pulp

and even infiltration of the trabecular arteries (Figs 1f and

1g). Conversely, in patients with partial or complete remis-

sion, BF, BV, and Ktrans increased paralleled by a reduction

of splenic volume. Compared to controls, BF, Ktrans, and

MTT were significantly different. Interestingly, negative

correlations between splenic volume and BF and BV were

present, while splenic volume was positively correlated with

MTT. For a further illumination of the effects during therapy,

we correlated the diameters of nodular lymphoma manifesta-

tions with the splenic perfusion parameters. Highly negative

correlations were present for BF (r = �0.777, P = .000) and

BV (r = �0.500, P = .011).

These results encourage a more extensive investigation of

splenic perfusion for staging and monitoring purposes in

patients with lymphoma. Of note, our small patient cohort

included mainly low-grade lymphoma and chronic leukemia,

which are known to elude diagnosis by FDG PET and are

therefore themost promising target group for further research.

Our study had some limitations. First, the small number of

patients with malignant hematologic disease involving the

spleen did not allow a comprehensive statistical evaluation.

Second, it is expected that the degree of tumor infiltration

of spleen parenchyma should be crucial for accurate detection,

with low-grade involvement probably eluding diagnosis.

Third, we had histologic proof of spleen involvement in

only one patient after splenectomy and correlation with

FDG PET in two patients. Especially in the lymphoma group,

the volume perfusion computed tomographic scan range did

not cover the entire spleen. Larger studies dealing with this

issue are necessary to establish the real benefit of using

VPCT in this clinical setting.

CONCLUSIONS

We demonstrated that VPCT can determine splenic perfusion

under both normal and pathologic conditions, offering the

possibility to analyze microcirculatory flow kinetics

throughout the spleen parenchyma. This information may

potentially discriminate between perfusion abnormalities

related to either portal hypertension or cell infiltration.

ACKNOWLEDGMENTS

We wish to thank our technicians, Nicole Sachse and Astrid

Schreiber, for their excellent assistance. Dr Leopoldine

Kotzina of the Institute of Pathology, Klinikum Sindelfingen-

B€oblingen, kindly prepared and analyzed the histologic spleensections from one of our patients.

8

REFERENCES

1. Ferrozzi F, Bova D, Draghi F, et al. CT findings in primary vascular tumors

of the spleen. AJR Am J Roentgenol 1996; 166:1097–1101.

2. De Schepper AM, Vanhoenacker F, Op de Beeck B, et al. Vascular

pathology of the spleen, part I. Abdom Imaging 2005; 30:96–104.

3. Bangerter M, Moog F, Buchmann I, et al. Whole-body 2-[18F]-fluoro-2-

deoxy-D-glucose positron emission tomography (FDG-PET) for accurate

staging of Hodgkin’s disease. Ann Oncol 1998; 9:1117–1122.

4. Castellino RA. Hodgkin disease: practical concepts for the diagnostic

radiologist. Radiology 1986; 159:305–310.

5. Pittaluga S, Verhoef G, Criel A, et al. ‘‘Small’’ B-cell non-Hodgkin’s

lymphomas with splenomegaly at presentation are either mantle cell

lymphoma or marginal zone cell lymphoma. A study based on histology,

cytology, immunohistochemistry, and cytogenetic analysis. Am J Surg

Pathol 1996; 20:211–223.

6. Mollejo M, Algara P, Mateo MS, et al. Large B-cell lymphoma presenting

in the spleen: identification of different clinicopathologic conditions. Am

J Surg Pathol 2003; 27:895–902.

7. Binet JL, Auquier A, Dighiero G, et al. A new prognostic classification of

chronic lymphocytic leukemia derived from amultivariate survival analysis.

Cancer 1981; 48:198–206.

8. Rai KR, Sawitsky A, Cronkite EP, et al. Clinical staging of chronic lympho-

cytic leukemia. Blood 1975; 46:219–234.

9. Swerdlow SH. WHO Classification of Tumours of Haematopoietic and

Lymphoid Tissues. Lyon, France: International Agency for Research on

Cancer, 2008.

10. UrbaWJ, Longo DL. Hodgkin’s disease. N Engl J Med 1992; 326:678–687.

11. Munker R, Stengel A, Stabler A, et al. Diagnostic accuracy of ultrasound

and computed tomography in the staging of Hodgkin’s disease. Verifica-

tion by laparotomy in 100 cases. Cancer 1995; 76:1460–1466.

12. Liu Y. Clinical significance of diffusely increased splenic uptake on

FDG-PET. Nucl Med Commun 2009; 30:763–769.

13. Dubois A, Dauzat M, Pignodel C, et al. Portal hypertension in lymphopro-

liferative and myeloproliferative disorders: hemodynamic and histological

correlations. Hepatology 1993; 17:246–250.

14. Treiber G, Csepregi A, Malfertheiner P. The pathophysiology of portal

hypertension. Dig Dis 2005; 23:6–10.

15. Tsushima Y, Unno Y, Koizumi J, et al. Measurement of human hepatic and

splenic perfusion using dynamic computed tomography: a preliminary

report. Comput Methods Programs Biomed 1998; 57:143–146.

16. MacMathuna P, O’Connor MK, Weir DG, et al. Non-invasive diagnosis of

portal vein occlusionby radionuclide angiography.Gut 1992; 33:1671–1674.

17. Martin-Comin J, Mora J, Figueras J, et al. Calculation of portal contribution

to hepatic blood flow with 99mTc-microcolloids. A noninvasive method to

diagnose liver graft rejection. J Nucl Med 1988; 29:1776–1780.

18. Peters AM,WalportMJ, Bell RN, et al. Methods ofmeasuring splenic blood

flow and platelet transit time with In-111-labeled platelets. J Nucl Med

1984; 25:86–90.

19. Ziegler SI, Haberkorn U, Byrne H, et al. Measurement of liver blood flow

using oxygen-15 labelled water and dynamic positron emission tomog-

raphy: limitations of model description. Eur J Nucl Med 1996; 23:169–177.

20. Miles KA, Charnsangavej C, Lee FT, et al. Application of CT in the investi-

gation of angiogenesis in oncology. Acad Radiol 2000; 7:840–850.

21. BlomleyMJ, Coulden R, Bufkin C, et al. Contrast bolus dynamic computed

tomography for the measurement of solid organ perfusion. Invest Radiol

1993; 28(Suppl 5):S72–S77.

22. Blomley MJ, Kormano M, Coulden R, et al. Splenic blood flow: evaluation

with computed tomography. Acad Radiol 1997; 4:13–20.

23. Nakashige A, Horiguchi J, Tamura A, et al. Quantitative measurement of

hepatic portal perfusion by multidetector row CT with compensation for

respiratory misregistration. Br J Radiol 2004; 77:728–734.

24. Goetti R, Leschka S, Desbiolles L, et al. Quantitative computed tomog-

raphy liver perfusion imaging using dynamic spiral scanning with variable

pitch: feasibility and initial results in patients with cancer metastases.

Invest Radiol 2010; 45:419–426.

25. Fujimoto K, Sawabe M, Sasaki M, et al. Undiagnosed cirrhosis occurs

frequently in the elderly and requires periodic follow ups and medical

treatments. Geriatr Gerontol Int 2008; 8:198–203.

26. Brown JJ, Naylor MJ, Yagan N. Imaging of hepatic cirrhosis. Radiology

1997; 202:1–16.

27. Chefd’hotel C, Hermosillo G, Faugeras O. Flows of diffeomorphisms

for multimodal image registration. In: Proceedings of the 2002 IEEE

International Symposium on Biomedical Engineering; 2002:753–756.

Academic Radiology, Vol -, No -, - 2012 SPLENIC PERFUSION

28. Saddi K, Chefd’hotel C, Cheriet F. Large deformation registration of

contrast-enhanced images with volume-preserving constraint. In: Pluim

J, Reinhardt J, eds. Medical Imaging 2007: Image Processing Bellingham,

WA: The International Society for Optical Engineering (SPIE); 2007.

29. Bruder HH, Flohr T, Raupach R. Method for improving the quality of

computed tomography image series by image processing and CT system

comprising a computational unit. Forchheim, Germany: Siemens Akeien-

gesellschaft, 2009.

30. Abbott RM, Levy AD, Aguilera NS, et al. From the archives of the AFIP:

primary vascular neoplasms of the spleen: radiologic-pathologic

correlation. Radiographics 2004; 24:1137–1163.

31. Mebius RE, Kraal G. Structure and function of the spleen. Nat Rev Immunol

2005; 5:606–616.

32. Groom AC, Schmidt EE, MacDonald IC. Microcirculatory pathways and

blood flow in spleen: new insights from washout kinetics, corrosion casts,

and quantitative intravital videomicroscopy. Scanning Microsc 1991; 5:

159–173.

33. Strijk SP, Boetes C, Bogman MJ, et al. The spleen in non-Hodgkin

lymphoma. Diagnostic value of computed tomography. Acta Radiol

1987; 28:139–144.

34. Daskalogiannaki M, Prassopoulos P, Katrinakis G, et al. Splenic involve-

ment in lymphomas. Evaluation on serial CT examinations. Acta Radiol

2001; 42:326–332.

35. de Jong PA, van Ufford HM, Baarslag HJ, et al. CT and 18F-FDG

PET for noninvasive detection of splenic involvement in patients

with malignant lymphoma. AJR Am J Roentgenol 2009; 192:

745–753.

36. Kazama T, Faria SC, Varavithya V, et al. FDG PET in the evaluation of

treatment for lymphoma: clinical usefulness and pitfalls. Radiographics

2005; 25:191–207.

9