Springer Science+Business Media, LLC 2010

Abdom Imaging (2010)

DOI: 10.1007/s00261-010-9608-6

Abdominal
Imaging

Bland and tumor thrombi in abdominal

malignancies: magnetic resonance imaging
assessment in a large oncologic patient
population
Marc Engelbrecht, Oguz Akin, Devesh Dixit, Lawrence Schwartz
Department of Radiology, Memorial Sloan-Kettering Cancer Center, New York, NY, USA

Abstract
The purpose of this study was to examine the distribution of venous thrombi associated with primary or secondary abdominal malignancies on magnetic resonance
(MR) imaging with respect to thrombus type (bland vs.
tumor), tumor sites, tumor types, and veins involved in a
large oncologic patient population. In a retrospective
review of 10,908 oncologic patients, MR imaging studies
identied 142 (1.3%) showing venous thrombi, of which
55 (0.5%) were bland and 87 (0.79%) were tumor
thrombus. Bland thrombi were most commonly seen in
liver (35%; 19/55) and retroperitoneal malignancies
(24%; 13/55) and were most often located in the inferior
vena cava (45%; 25/55) and the portal vein (22%; 12/55).
Tumor thrombi were most commonly seen in renal (55%;
48/87) and liver (32%; 28/87) malignancies. The prevalence of tumor thrombi was 8.8% (48/545) in primary
renal, 4.7% (6/126) in primary retroperitoneal, 2.9% (19/
634) in primary liver, and 1.8% (9/479) in secondary liver
malignancies. Tumor thrombi were most commonly located in the inferior vena cava (57%; 50/87), the renal
vein (48%; 42/87), and the portal vein (29%; 25/87).
Key words: MRIThrombusBland
thrombusTumor thrombusOncology

Both bland and tumor thrombi represent important

complications and therapeutic challenges in oncologic
patients. Ultrasound, computed tomography (CT), or
magnetic resonance (MR) imaging can provide valuable
information for the management of oncologic patients
with thrombosis. The main advantage of MR imaging is

Correspondence to: Oguz Akin; email: akino@mskcc.org

its superior soft tissue contrast resolution, which allows

improved characterization of thrombus composition.
The purpose of our study was to examine the distribution
of bland and tumor thrombi on MR imaging studies of
patients with abdominal primary or secondary malignancies with respect to tumor sites, types, and veins involved.

Materials and methods

Our institutional review board approved and issued a
waiver of informed consent for our retrospective study,
which was compliant with the Health Insurance Portability and Accountability Act.

Patients
By searching an institutional database, we identied a
total of 10,908 patients treated at our institution (a
dedicated cancer center) for primary or secondary
malignancies in the abdomen who underwent MR
imaging in our Radiology Department in a period of
10 years. The reports for these MR imaging studies were
then searched electronically for the following terms:
thrombus, thrombi, thrombosis, thrombosed, clot, clots,
clotted, embolus, emboli, embolized, thromboembolus, and
thromboemboli.
A total of 670 MR imaging studies contained one of
the keywords. After manual review of these reports, 471
studies were excluded because they indicated the absence
of thrombus. In addition, 28 studies were excluded because they were follow-up exams for patients with
thrombus in the same vessel imaged in an earlier MR
imaging study. Thus, a total of 171 MR imaging studies
in 170 patients (106 male, mean age: 57; 64 female, mean
age: 59) were included in our study. One patient had a
thrombus on follow-up MR imaging in a different vessel.

M. Engelbrecht et al.: Bland and tumor thrombi in abdominal malignancies

Table 1. Bland thrombus distribution in the abdomen based on tumor sites, tumor types, and veins involved

Liver
Liver metastasisa
Hepatocellular carcinoma
Cholangiocarcinoma
Hepatoblastoma
Retroperitoneum
Retroperitoneal sarcoma
Germ cell tumor
Lymphoma
Pancreas
Pancreatic cancer
Kidney
Renal cell carcinoma
Urothelial cancer
Wilms tumor
Adrenal
Neuroblastoma
Adrenocortical carcinoma
Pheochromocytoma
Unknown primary
Total

No.

IVC

RV

PV

HV

SMV

SV

19
14
3
1
1
13
2
7
4
9
9
8
3
4
1
5
3
1
1
1
55

5
4

4
1
3

8
7

1
1

1
1

1
1

2
2

1
10
4
3
3
1
1
6
2
3
1
3
2
1
25

1
1

3
3

5
5

4
4

2
1
1
1

12

1
1
3

1
1

Liver metastases originated from colon cancer (n = 5), breast cancer (n = 1), pancreatic cancer (n = 1), adrenocortical carcinoma (n = 1),
cholangiocarcinoma (n = 1), unknown primary (n = 1), prostate cancer (n = 1), ovarian cancer (n = 1), renal cell carcinoma (n = 1), and lung
cancer (n = 1)
IVC, inferior vena cava (including proximal common iliac veins); RV, renal vein; PV, portal vein; HV, hepatic vein; SMV, superior mesenteric vein;
SV, splenic vein

MR imaging examinations
All MR imaging studies were performed at our institution on a 1.5-T whole-body MR imager (GE Medical
Systems, Milwaukee, WI, USA) and were entered into
our picture archiving and communication system
(PACS). Although during the study period MR imaging parameters changed and MR imaging protocols
varied depending on the type of malignancy and clinical question asked, all MR imaging examinations
contained at least the following sequences: in-phase
and opposed-phase T1-weighted gradient-echo imaging;
fat-saturated fast spin-echo T2-weighted imaging; single-shot fast spin-echo T2-weighted imaging; and dynamic multiphase gadolinium-enhanced T1-weighted
gradient-echo imaging. Additional steady-state free
precession sequences were also included in most protocols.

Data collection
Demographic and clinical information for each patient
were obtained from the electronic medical records. The
ofcial radiology reports of 171 MR imaging studies
were used to classify the thrombus type as tumor or
bland and to determine the location and extent of each
tumor thrombus. A board-certied fellowship-trained
attending radiologist with expertise (4 years of experience post-body imaging fellowship) in oncologic body
imaging and MR imaging retrospectively reviewed all
these MR imaging studies to help ensure consistency in

categorizing the thrombus type, location, and extent. The

main feature used to differentiate tumor thrombus from
bland thrombus was the presence or absence of
enhancement. Additional ndings such as continuity of
tumor with intravascular thrombus, vessel expansion by
thrombus, and thrombus signal intensity characteristics
similar to those of adjacent tumor were also used to
identify tumor thrombus. Cases of mixed tumor and
bland thrombus within the same vessel were recorded as
tumor thrombus.

Results
Overall, we identied venous thrombi in 1.3% (142/
10,908) of oncologic MR imaging studies: 55 (0.5%) were
bland thrombi and 87 (0.8%) were tumor thrombi.

Bland thrombi distribution

Bland thrombi (Table 1; Fig. 1) were most commonly
associated with liver (35%; 19/55) and retroperitoneal
malignancies (24%; 13/55). The majority (74%; 14/19) of
bland thrombi in liver malignancies were associated with
liver metastases.
Bland thrombi most commonly involved the inferior
vena cava (45%; 25/55) and the portal vein (22%; 12/55).
Retroperitoneal malignancies were the predominant
cause of bland thrombi in the inferior vena cava,
accounting for 40% (10/25). Liver malignancies were the
predominant cause of bland thrombi in the portal vein,
accounting for 67% (8/12).

M. Engelbrecht et al.: Bland and tumor thrombi in abdominal malignancies

Discussion

Fig. 1. Fifty-two-year-old female after resection of a right

adrenal carcinoma. Transverse fat-saturated T2-weighted MR
image (A) shows a small thrombus (arrow) in the inferior vena
cava. Coronal post-contrast fat-saturated T1-weighted MR
image (B) shows no enhancement of the thrombus, indicating
that it is a bland thrombus (arrow).

Distribution of tumor thrombi

Tumor thrombi (Table 2; Figs. 2, 3, 4, and 5) were most
commonly due to renal (55%; 48/87) and liver malignancies (32%; 28/87). Overall, tumor thrombi were
present in 8.8% (48/545) of renal malignancies, 2.9% (19/
634) of primary liver malignancies, 1.8% (9/479) of secondary liver malignancies, and 4.7% (6/126) of retroperitoneal sarcomas (Table 3).
Tumor thrombi most commonly involved the inferior vena cava (57%; 50/87), the renal vein (48%; 42/87),
and the portal vein (29%; 25/87). Seventy percent (35/
50) of tumor thrombi in the inferior vena cava and 93%
(39/42) of tumor thrombi in the renal vein were due to
renal malignancies. Ninety-two percent (23/25) of
tumor thrombi in the portal vein were due to liver
malignancies.

To the best of our knowledge, our series is the largest yet

examined to determine the prevalence of bland and tumor thrombi on MR imaging studies in patients with
abdominal primary or secondary malignancies. Our retrospective review of 10,908 MR imaging studies found
that, although they were rare pathologies, bland thrombi
were most commonly seen in liver and retroperitoneal
malignancies, whereas tumor thrombi were most commonly seen in renal and liver malignancies. Bland
thrombi most commonly involved the inferior vena cava
and the portal vein, and tumor thrombi most commonly
involved the inferior vena cava, the renal vein, or the
portal vein.
The increased risk of bland venous thrombosis in
cancer patients is well known. Although malignancy itself is an important risk factor for venous thrombosis,
other contributing risk factors in these patients include
concurrent chemotherapy, recent surgery, neurologic
disease, and immobility. Tumor thrombus is also a very
important complication in oncologic patients and results
from either direct extension of an adjacent malignancy
or, more rarely, arises as a primary malignancy of the
involved vessel.
Both CT and MR imaging play important roles in
detecting, characterizing, and determining the extent of
thrombi in patients with cancer. Although CT is used more
often than MR imaging in these patients, MR imaging
offers some advantages in the detection and characterization of thrombi, such as the possibility of omitting
contrast and using ow-sensitive sequences for detection.
The superior soft tissue contrast resolution of MR imaging
is very useful for characterizing thrombus composition.
Because of the T2 shortening effect of blood breakdown
products, bland thrombi are distinctly low in signal
intensity compared with tumor thrombi, which are of
intermediate signal intensity. The enhancement that occurs in tumor thrombi is an important nding for differentiating such thrombi from bland thrombi on CT as well
as MR imaging. However, this enhancement is usually
more conspicuous on MR imaging.
When a vascular lling defect is seen in an oncologic
patient on MR imaging, the rst step is to rule out a
malignant cause [1]. Direct extension of tumor into the
vessel lumen, expansion of the vessel by the tumor, heterogeneous and increased thrombus signal intensity on
T2-weighted MR images, and thrombus enhancement
are characteristics of tumor thrombi [14]. Collateral
vessels around a bland thrombus, such as cavernous
transformation in portal vein thrombosis, should not be
confused with enhancement of the thrombus itself.
Artifactual filling defects resulting from the mixture of
opacified and unopacified blood may occasionally mimic
true filling defects, but delayed post-contrast images
confirm their artifactual nature.

M. Engelbrecht et al.: Bland and tumor thrombi in abdominal malignancies

Table 2. Tumor thrombus distribution in the abdomen based on tumor sites, tumor types, and veins involved

Kidney
Renal cell carcinoma
Urothelial cancer
Renal sarcoma
Wilms tumor
Liver
Hepatocellular carcinoma
Liver metastasisa
Cholangiocarcinoma
Retroperitoneum
Retroperitoneal sarcoma
Germ cell tumor
Lymphoma
Others
Carcinoid tumor
Recurrent pancreatic cancer
Adrenal metastasisb
Total

No.

IVC

RV

48
42
3
2
1
28
15
9
4
8
6
1
1
3
1
1
1
87

35
32
2
1

39
34
3
1
1

8
4
3
1
6
4
1
1
1

3
3

PV

HV

SMV

SV

2
2

2
2

GV

2
2

23
13
6
4
1
1

7
3
4

1
1

1
1

1
1

1
50

42

25

a
Liver metastases originated from colon cancer (n = 5), melanoma (n = 1), germ cell tumor (n = 1), renal cell carcinoma (n = 1), and unknown
primary (n = 1)
b
Adrenal metastasis originated from lung cancer (n = 1)
IVC, inferior vena cava (including proximal common iliac veins); RV, renal vein; PV, portal vein; HV, hepatic vein; SMV, superior mesenteric vein;
SV, splenic vein, GV, gonadal vein

Fig. 2. Eighty-two-year-old male with renal cancer. Transverse (A) and coronal (B) T2-weighted MR images show a right
renal mass (M) with tumor thrombus (arrows) extending into the

right renal vein and the inferior vena cava. Note the enhancement of the mass (M) and the tumor thrombus (arrows) on postcontrast fat-saturated T1-weighted MR image (C).

Fig. 3. Eighty-two-year-old male with hepatocellular carcinoma. Transverse T2-weighted fat-saturated MR image (A)
shows a liver mass (M) with tumor thrombus (arrows) in the
left portal vein, which is expanded by the tumor thrombus.

Pre- (B) and post-contrast (C) fat-saturated T1-weighted MR

images again demonstrate the mass (M) and the tumor
thrombus (arrows). Note the enhancement of the mass (M)
and the tumor thrombus (arrows) (C).

M. Engelbrecht et al.: Bland and tumor thrombi in abdominal malignancies

Fig. 4. Fifty-two-year-old male with intrahepatic cholangiocarcinoma. Transverse fat-saturated T2-weighted (A) and
post-contrast fat-saturated T1-weighted (B) MR images
show a large liver mass (M) with tumor thrombus (arrows)

Determining the extent of a tumor thrombus is

important for treatment planning. It has been reported
that complete surgical resection of tumor thrombus improves survival in renal cancer [58]. Furthermore, the
surgical approach and scope are determined by the level
of extension of tumor thrombus. For example, tumor
thrombus in the inferior vena cava extending into the
right atrium would require a thoracoabdominal approach and cardiopulmonary bypass during renal cancer
surgery [9]. Therefore, preoperative assessment for the
presence and the extent of renal vein and inferior vena
cava tumor thrombi is important for planning subsequent treatment and choosing the appropriate surgical
approach. Both CT and MR imaging are very effective in
detecting the presence of tumor thrombi, delineating
their extent, and differentiating them from bland
thrombi in renal malignancies [1016].
Likewise, tumor thrombus in the portal vein is an
important complication and a prognostic factor in

involving the hepatic vein and the inferior vena cava. Coronal T2-weighted MR image (C) shows the liver mass (M) and
the tumor thrombus (arrows) involving the inferior vena cava
and extending superiorly into the right atrium.

hepatocellular carcinoma [1719]. It often leads to

extensive spreading of the tumor throughout the liver
and can increase portal venous pressure. Portal venous
invasion in patients with hepatocellular carcinoma is a
major therapeutic challenge and, when left untreated, is
associated with reduced survival. Among a number of
treatments assessed, combined treatments consisting of
hepatectomy and transcatheter arterial chemoembolization, chemotherapy, or internal radiation have yielded
the most promising results [20]. In addition to primary
hepatocellular carcinoma, it has been reported that
metastases can also result in tumor thrombus in the liver
[21, 22].
We acknowledge the limitations of our study. First,
our results may have been biased by the selection of only
MR imaging studies for analysis. For example, it may
well be that only patients in whom the presence of
thrombosis was suspected but uncertain underwent MR
imaging, while patients with no thrombosis or very

M. Engelbrecht et al.: Bland and tumor thrombi in abdominal malignancies

Table 3. Prevalence of tumor thrombi in most commonly associated

malignancies
Type of malignancy

Renal
Primary liver
Secondary liver
Retroperitoneal sarcoma

Total no.
of patients
with malignancy

545
634
479
126

No. and
percentage of
patients with
associated tumor
thrombi
48
19
9
6

(8.8%)
(2.4%)
(1.8%)
(4.7%)

however, we did find a rate of tumor thrombus (8.8%)

within the range reported in the literature (410%) [5, 23
25]. Furthermore, the retrospective nature and the long
duration of the study resulted in heterogeneity in the
techniques and parameters used for MR imaging. Finally, because of the large number of patients examined
during the study period, eligible patients had to be
identified through a computerized search of original MR
imaging reports rather than a retrospective review of the
MR imaging studies themselves.
In conclusion, we showed the distribution and prevalence of bland and tumor thrombi in abdominal
malignancies assessed by MR imaging in a large oncologic patient population. Our results can benet radiologists by alerting them as to when, and in which veins,
they may expect thrombus when reading oncologic
abdominal MR imaging studies.
References

Fig. 5. Seventy-three-year-old female with retroperitoneal

sarcoma. Transverse fat-saturated T2-weighted (A) and postcontrast fat-saturated T1-weighted (B) MR images show a
large heterogeneous retroperitoneal mass (M) with tumor
thrombus (arrow) involving the inferior vena cava. Note that
the mass (M) abuts the aorta (open arrow) but does not invade it (A, B).

obvious thrombosis underwent CT, preventing accurate

estimation of the true prevalence of thrombus in our
patient group. For example, based predominantly on
CT, another study reported a much higher prevalence of
portal vein thrombosis (31%) among patients with
hepatocellular carcinoma [17]. In renal cell cancer,

1. Kaufman LB, Yeh BM, Breiman RS, et al. (2005) Inferior vena
cava filling defects on CT and MRI. AJR Am J Roentgenol
185:717726
2. Tublin ME, Dodd GD 3rd, Baron RL (1997) Benign and malignant
portal vein thrombosis: differentiation by CT characteristics. AJR
Am J Roentgenol 168:719723
3. Aslam Sohaib SA, Teh J, Nargund VH, et al. (2002) Assessment of
tumor invasion of the vena caval wall in renal cell carcinoma cases
by magnetic resonance imaging. J Urol 167:12711275
4. Ergen FB, Hussain HK, Caoili EM, et al. (2004) MRI for preoperative staging of renal cell carcinoma using the 1997 TNM classification: comparison with surgical and pathologic staging. AJR
Am J Roentgenol 182:217225
5. Hatcher PA, Anderson EE, Paulson DF, Carson CC, Robertson JE
(1991) Surgical management and prognosis of renal cell carcinoma
invading the vena cava. J Urol 145:2023
6. Skinner DG, Pritchett TR, Lieskovsky G, Boyd SD, Stiles QR
(1989) Vena caval involvement by renal cell carcinoma. Surgical
resection provides meaningful long-term survival. Ann Surg
210:387392
7. Neves RJ, Zincke H (1987) Surgical treatment of renal cancer with
vena cava extension. Br J Urol 59:390395
8. ODonohoe MK, Flanagan F, Fitzpatrick JM, Smith JM (1987)
Surgical approach to inferior vena caval extension of renal carcinoma. Br J Urol 60:492496
9. Tsuji Y, Goto A, Hara I, et al. (2001) Renal cell carcinoma with
extension of tumor thrombus into the vena cava: surgical strategy
and prognosis. J Vasc Surg 33:789796
10. Johnson CD, Dunnick NR, Cohan RH, Illescas FF (1987) Renal
adenocarcinoma: CT staging of 100 tumors. AJR Am J Roentgenol
148:5963
11. Horan JJ, Robertson CN, Choyke PL, et al. (1989) The detection of
renal carcinoma extension into the renal vein and inferior vena

M. Engelbrecht et al.: Bland and tumor thrombi in abdominal malignancies

12.
13.

14.
15.
16.
17.
18.

cava: a prospective comparison of venacavography and magnetic

resonance imaging. J Urol 142:943948
Myneni L, Hricak H, Carroll PR (1991) Magnetic resonance
imaging of RCC with extension into the vena cava: staging accuracy and recent advances. Br J Urol 68:571578
Kallman DA, King BF, Hattery RR, et al. (1992) Renal vein and
inferior vena cava tumor thrombus in renal cell carcinoma: CT,
US, MRI and venacavography. J Comput Assist Tomogr 16:240
247
Roubidoux MA, Dunnick NR, Sostman HD, Leder RA (1992)
Renal cell carcinoma: detection of venous extension with gradientecho MR imaging. Radiology 182:269272
Welch TJ, LeRoy AJ (1997) Helical and electron beam CT scanning
in the evaluation of renal vein involvement in patients with renal
cell carcinoma. J Comput Assist Tomogr 21:467471
Laissy JP, Menegazzo D, Debray MP, et al. (2000) Renal carcinoma: diagnosis of venous invasion with Gd-enhanced MR
venography. Eur Radiol 10:11381143
Connolly GC, Chen R, Hyrien O, et al. (2008) Incidence, risk
factors and consequences of portal vein and systemic thromboses in
hepatocellular carcinoma. Thromb Res 122:299306
Jonas S, Bechstein WO, Steinmuller T, et al. (2001) Vascular
invasion and histopathologic grading determine outcome after liver
transplantation for hepatocellular carcinoma in cirrhosis. Hepatology 33:10801086

19. Takizawa D, Kakizaki S, Sohara N, et al. (2007) Hepatocellular

carcinoma with portal vein tumor thrombosis: clinical characteristics, prognosis, and patient survival analysis. Dig Dis Sci 52:3290
3295
20. Minagawa M, Makuuchi M (2006) Treatment of hepatocellular
carcinoma accompanied by portal vein tumor thrombus. World J
Gastroenterol 12:75617567
21. Heaston DK, Chuang VP, Wallace S, de Santos LA (1981) Metastatic hepatic neoplasms: angiographic features of portal vein
involvement. AJR Am J Roentgenol 136:897900
22. Lee KF, Chu W, Lai PB (2005) Portal vein tumor thrombus in
colorectal liver metastasis. Am J Surg 190:364365
23. Hoehn W, Hermanek P (1983) Invasion of veins in renal cell carcinomafrequency, correlation, and prognosis. Eur Urol 9:276
280
24. Nesbitt JC, Soltero ER, Dinney CPN, et al. (1997) Surgical management of renal cell carcinoma with inferior vena cava tumor
thrombus. Ann Thorac Surg 63:15921600
25. Pagano F, Dal Bianco M, Artibani W, et al. (1992) Renal cell
carcinoma with extension into the inferior vena cava: problems in
diagnosis, staging, and treatment. Eur Urol 22:200203