Angiography effectively supports the diagnosis of hepatic metastases in medullary thyroid carcinoma

Angiography Effectively Supports the Diagnosis ofHepatic Metastases in Medullary Thyroid Carcinoma

Olga Esik, M.D., D.Sc.1

Peter Szavcsur, M.D.2

Szabolcs Szakall, Jr., M.D.3

Gabor Bajzik, M.D.4

Imre Repa, M.D., Ph.D.4

Gabriella Dabasi, M.D., Ph.D.5

Marton Fuzy, M.D., Ph.D.6

Zoltan Szentirmay, M.D., Ph.D.7

Ferenc Perner, M.D., D.Sc.5

Miklos Kasler, M.D., Ph.D.8,9

Zsolt Lengyel, M.D.3

Lajos Tron, Ph.D., D.Sc.3

1 Department of Radiotherapy, Semmelweis University,National Institute of Oncology, Budapest, Hungary.

2 Department of Diagnostic Imaging, National In-stitute of Oncology, Budapest, Hungary.

3 Positron Emission Tomography Center, Universityof Debrecen, Debrecen, Hungary.

4 Department of Diagnostic Imaging, University ofKaposvar, Kaposvar, Hungary.

5 Department of Transplantation and Surgery, in-cluding Section of Nuclear Medicine, SemmelweisUniversity, Budapest, Hungary.

6 Department of Nuclear Medicine, National Insti-tute of Oncology, Budapest, Hungary.

7 Department of Molecular Pathology, National In-stitute of Oncology, Budapest, Hungary.

8 Department of Head and Neck Surgery, NationalInstitute of Oncology, Budapest, Hungary.

9 Department of Oncology, Semmelweis Univer-sity, Budapest, Hungary.

Supported in part by grants Scientific Council ofHealth Care (ETT) 12/1996, and National ScientificResearch Fund (OTKA) T25827 and T29809.

Address for reprints: Olga Esik, M.D., D.Sc., De-partment of Radiotherapy, Semmelweis Universityand National Institute of Oncology, Rath Gyorgy u.7-9, H-1122 Budapest, Hungary; Fax: 36-1-2248620; E-mail: [email protected]

Received May 16, 2000; revisions received No-vember 8, 2000 and January 23, 2001; acceptedJanuary 26, 2001.

BACKGROUND. Medullary thyroid carcinoma (MTC) belongs in the group of neu-

roendocrine tumors with early lymphatic and hepatic dissemination. A high rate of

undetectable metastases is hypothesized to be responsible for the frequent mis-

match between the apparent relatively small tumor burden and the elevated

plasma tumor marker level.

METHODS. Thirty-six MTC patients with residual/recurrent biochemical signs (el-

evated basal calcitonin level) and/or characteristic general symptoms (diarrhea

and/or flushing) were systematically examined by conventional radiology, whole-

body 18F-deoxyglucose positron emission tomography (PET), dynamic liver com-

puted tomography and magnetic resonance imaging, and hepatic angiography.

RESULTS. Conventional diagnostic imaging revealed lymph node (LN) involvement

in the cervical, mediastinal, supraclavicular, and axillary regions (16 cases), and

multiple pulmonary (3 cases), bony (1 solitary and 1 multiple case), and breast (1

case) metastases. 18F-deoxyglucose PET identified all these extralymphatic meta-

static lesions (except 2 cases with multiple pulmonary metastases), and also

supradiaphragmatic LN involvement in 34 (94%) patients. In 32 (89%) cases,

multiple small (generally # 1 cm) hypervascular, hepatic metastases undetectable

by other imaging methods were localized angiographically. Of the 23 original

pathologic specimens investigated, 18 (78%) exhibited LN involvement. The small-

est primary tumor in patients with hepatic metastases was 1 cm.

CONCLUSIONS. Hepatic angiography is recommended for primary staging in MTC

patients with a primary tumor measuring 1 cm or larger, and/or pathologically

proven LN involvement, and also during restaging for suspected recurrences to

avoid unnecessary extensive surgical LN dissection in the neck and mediastinum.

Cancer 2001;91:2084 –95. © 2001 American Cancer Society.

KEYWORDS: medullary thyroid carcinoma, hypercalcitoninemia, hepatic metastasis,angiography.

Medullary thyroid carcinoma (MTC) is a rare, malignant neuroen-docrine tumor originating from the parafollicular C cells of the

thyroid gland. The clinical course may be capricious but is generallyindolent and frequently involves early lymphatic dissemination.Pathologic examinations of the lymph nodes (LNs) during primarystaging reveals cervical LN involvement in 31–33% of the pT1 cases,1,2

in 53% of the pT2 cases,2 and in 100% of the pT3– 4 cases,2 withintermediate data for nonselected patient groups.3–7 The incidence ofprimary mediastinal LN involvement has been reported to be 22–50%,2,7,8 as a function of the patient selection. Distant metastases areobserved in only 2–33% of the cases during the primary staging,depending on the form (an inherited vs. a noninherited form, the typeof inheritance, and an early vs. an advanced stage) of the dis-

2084

© 2001 American Cancer Society

ease.5,6,9,10 The incidence of distant metastases is gen-erally higher (45%) during the follow-up as the diseaseprogresses.6

After the completion of primary treatment, persis-tent or recurrent hypercalcitoninemia is reported tooccur in 29 – 85% of the cases.1,2,8,9,11,12 Localizationattempts after the observation of increased plasmatumor marker levels often fail to reveal marker-pro-ducing foci, and secondary surgical interventions afterthe detection of presumably marker-producing tu-morous foci rarely (0 –38%) achieve a biochemicalcure.2,8,11–20 These data conflict with the publishedrelatively low incidence of distant metastases in MTCpatients and suggest that subdetectable tumorous fociare responsible for the unexplained elevated plasmatumor marker levels in these patients. The currentcommunication describes a high incidence of hepaticmetastases in a cohort of MTC patients with a mis-match between high tumor marker levels and/or gen-eral symptoms (diarrhea/flushing) characteristic of re-currence and small detected tumor masses whenconventional diagnostic methods are used.

METHODSPatientsThe study was conducted between June 1998 and De-cember 1999 and included 36 MTC patients (14 males,22 females) meeting the inclusion criteria, but other-wise consecutive in our consulting program. Some ofthe patients were on first referral, whereas others afterthe primary diagnostic workup and primary surgicalintervention were participants in our regular follow-up program. The mean time between the first sign ofthe residual/recurrent disease and the study was 45months (range, 0 –232 months), in accordance withthe relatively long MTC history of the patients. Inclu-sion criteria comprised an elevated plasma tumormarker (calcitonin) level (34 cases) and/or (11 cases)characteristic general symptoms (diarrhea and/orflushing).

The mean age at primary MTC diagnosis was 46years (range, 22– 66 years). Clinically, the numbers ofmultiple endocrine neoplasia (MEN)2a, familial non-MEN, and undetermined (sporadic or hereditary)background MTC cases in the cohort were 11, 6, and19, respectively. The primary thyroid surgery included20 near-total thyroidectomies (12 cases in 1 step, 7cases in 2 steps, and 1 case in 3 steps), 12 bilateralsubtotal thyroidectomies (9 cases in 1 step and 3 casesin 2 steps), 1 unilateral subtotal thyroidectomy, 2 me-diastinally located MTC extirpations, and 1 case inwhich the thyroidectomy was performed only 14 yearsafter the primary care (the primary surgical interven-tion was merely cervical LN dissection). During the

primary surgical treatment, 23 patients underwentsome kind of LN dissection (21 from the cervical com-partments and 2 from the upper mediastinum), theextent of which ranged from lymph node picking toradical dissections. During the seven secondary surgi-cal interventions, one patient was operated on forbilateral subtotal thyroidectomy, and six patients fordifferent types of LN dissection (including two caseswho underwent near-total thyroidectomy). Two pa-tients participated in tertiary surgical care (one fortotal thyroidectomy and another one for cervical LNdissection).

Twenty-eight patients received external irradia-tion; the average dose of the 27 cervical irradiationsamounted to 45.7 grays (Gy), whereas that of the 10mediastinal treatments was 44.3 Gy. 131I-meta-iodo-benzylguanidine (MIBG) radionuclide therapy wasprovided in 13 cases, with a mean dose of 5278 MBq.After a primary misdiagnosis, 1 subject received131I-Na therapy of 6543 MBq. All cases were patholog-ically reevaluated, and the MTC assignment was sup-ported by immunophenotyping.

Imaging MethodsAll patients regularly participated in follow-up imag-ing: cervical and abdominal ultrasonography (US),cervical, mediastinal, and upper abdominal computedtomography (CT) and magnetic resonance imaging(MRI), 99mTc bone scintigraphy (additional CT andMRI for suspected lesions), and whole-body 131I-MIBGscintigraphy (diagnostic or posttherapeutic). Soma-tostatin receptor (SR) scintigraphy (SRS) with 111In-octreotide was performed in two patients only, in con-sequence of the inaccessibility of this examination forroutine use. Because these investigations were per-formed in different diagnostic departments with dif-ferent equipment and methodology, they were notreevaluated in the current study.

Each patient was examined by 18F-deoxyglucose(FDG) positron emission tomography (PET), dynamicliver CT and MRI, and hepatic angiography in the18-month study period. The four examinations of eachpatient were completed within 4 months, a periodcomparing favorably with the generally slow progres-sion of MTC. The investigations were separated by 2months or more from the end of the previous thera-peutic intervention. Written consent was obtained be-fore the diagnostic procedures were performed.

Whole-body FDG PET studies were performed af-ter a 4-hour fasting by using a GE 4096 Plus PETscanner (General Electric, Uppsala, Sweden) providing15 two-dimensional sections over an axial field of view(FOV) of 103 mm. The in-plane resolution was 5.5 mm,and the axial resolution was 6 mm. Whole-body ex-

Hepatic Metastases in Medullary Thyroid Carcinoma/Esik et al. 2085

aminations started from the base of the pelvis, 40minutes after the administration of 5.55 MBq/kg FDG.The acquired data were reconstructed by a filteredback-projection (4.2-mm Hanning filter) without at-tenuation correction. To determine the number ofinvolved LNs, we supplemented visual interpretationof the reconstructed images by a lesion-to-back-ground ratio calculation, using regions-of-interest de-fined for each lesion with nonphysiologic tracer accu-mulation in locations characteristic of the lymphaticdrainage of the thyroid gland (central cervical, left andright lateral cervical, mediastinal, left and right supra-clavicular and axillary regions), and for the back-ground (muscles).21 Lymph nodes with lesion-to-background ratios of 2 or more were regarded asmetastatic. Two specialists retrospectively read thePET images of all patients. The findings were com-pared and in cases of discordance a consensus wasreached.

Hepatic foci were investigated by dynamic CT andMRI, and hepatic angiography. All MRI examinations(35 patients, 1 claustrophobic case omitted) were con-ducted on a 1.5-T system (Magnetom Vision; SiemensMedical Systems, Inc., South Iselin, NJ) with a bodycoil; in 22 cases, a phased-array body coil was used.First, T2-weighted turbo spin-echo sequences withchemical shift fat saturation, breath-held T2-weightedturbo spin-echo sequences, and T1-weighted fast low-angle shot (FLASH) sequences without fat saturationwere obtained. In the next step, a dynamic gadolini-um-enhanced study followed. Finally, T1-weightedbreath-held FLASH sagittal and/or coronal sequenceswere acquired.

Axial fat-suppressed T2-weighted turbo spin-echo(TR/TE 5 4000/83 and 165, echo train length 9) imag-ing was performed with 4-signal acquisition (1083 256 matrix, FOV varying with patient size: 350 – 420mm, with a 5/8 rectangular FOV). During a 20-secondbreath-hold, axial T2-weighted turbo spin-echo im-ages were obtained (TR/TE 5 3800/138, echo trainlength 29, 116 3 256 matrix, FOV 350 – 420 mm, with a5/8 rectangular FOV, 1 signal acquired). The axial T1-weighted breath-held FLASH imaging involved the fol-lowing parameters: TR/TE 5 273/4.8, 70° flip angle,102 3 256 matrix, FOV 350 – 420 mm, with a 5/8 rect-angular FOV, 1 signal acquired. The dynamic studycomprised 13 axial sections acquired 6 times every 40seconds after a rapid bolus injection of gadodiamid(Omniscan, Nycomed, Oslo, Norway; 0.1 mmol per kgbody weight, followed by 20 mL saline) during a 28-second breath-hold. All sequences were performedwith a 10-mm slice thickness and a 25% interslice gap.Cephalic and caudal presaturation pulses were ap-plied to minimize motion artefacts. The detection of

the lesions was enhanced by image subtraction. Forthe T1-weighted breath-held FLASH coronal and sag-ittal imaging, the same sequence was used as in axialimaging.

All CT examinations (36 patients) were performedon Somatom Plus 4 (Siemens) equipment. First, un-enhanced helical scans were obtained with 8 mm col-limation and a pitch of 1:1. The entire liver was cov-ered in 20 –26 seconds. Two other scans were obtainedwith the same parameters, 24 and 65 seconds after thestart of a bolus administration of intravenous contrastmaterial via an injector (iopamidole, Iopamiro;Bracco, Milan, Italy; 1.5 mL per kg body weight, con-taining 370 mg iodine/mL, at 3 mL/second). Eachseries was obtained during a single breath-hold, at 120kV and 220 mAs. The reconstruction interval was 8mm for the unenhanced and 6 mm for the arterial andportal phase images.

Two radiologists retrospectively read the CT andMRI images of all patients. The findings were com-pared, and in cases of discordance a consensus wasreached. Liver lesions were divided into three catego-ries: hypervascularized lesions, cysts, and nondeter-minable lesions.22–26 The hypervascularized lesionswere hypodense or isodense to the liver parenchymabefore contrast material administration. They werepartly or completely hyperdense in the arterial phase,and isodense or hypodense in the portal phase. Theselesions furnished low signal intensity in T1-weightedimages, and high signal intensity in T2-weighted im-ages, and their contrast enhancement pattern wassimilar to that in the CT examination. Cysts were oflow density in CT (0 –15 Hounsfield [HU]), low signalintensity in T1-weighted images, and high signal in-tensity in T2-weighted images and showed no en-hancement. Nondeterminable lesions were hypo- orisodense in CT before contrast medium administra-tion and displayed some enhancement but remainedhypodense in the arterial and portal phases as com-pared with with the liver parenchyma. They yieldedlow signal intensity in T1-weighted images, produceda moderate delayed contrast enhancement during thedynamic study, and usually were undetectable orslightly hyperintense on T2-weighted images.

Hepatic angiography was performed after punc-ture of the the right femoral artery (Seldinger method)with a SIM 3 or Cobra 3 catheter. Results were evalu-ated with Digitron equipment (Siemens). First celia-cography, and then selective a. hepatica propriaangiography were performed with Omnipaque (Ny-comed) 350 contrast material (0.3 mL per kg bodyweight at 4 mL/second). Due to anatomic variations,the a. hepatica accessoria dextra and/or a. hepaticaaccessoria sinistra also were sometimes cannulated.

2086 CANCER June 1, 2001 / Volume 91 / Number 11

All hepatic angiograms were reread for this study bythe original examiner. Hepatic lesions detected by an-giography were taken as metastatic if they resembled ahypervascular focus and were clearly seen and distin-guished from the surrounding hepatic tissue in boththe arterial and parenchymal phases. Within foci mea-suring greater than 7 mm, lacunar contrast materialdistribution facilitated the diagnosis of metastases.

Laparoscopy of the liver was performed followingangiography in five patients, during laparoscopic cho-lecystectomy in two cases, and with the purpose ofliver diagnostics in three cases.

Tumor Marker DeterminationThe basal plasma calcitonin level was measured byELSA human calcitonin assay (CIS bio international,Gif sur Yvette, France) with two monoclonal antibod-ies, and carcinoembryonic antigen (CEA) was deter-mined by luminometry. The upper limit of the normalrange was set at 10 pg/mL for calcitonin, and 6 ng/mLfor CEA. Calcitonin and CEA levels were assessed 2months after surgery27 and later at 6-month intervals.Recurrent plasma tumor marker elevations and gen-eral symptoms suggest a previous 6-month markernegative or symptom free period.

RESULTSTable 1 shows pathologic data, marker levels, andresults of imaging methods. The mean primary tumorsize was 3.3 cm (range, 1.0 –7.0 cm). Lymph nodeswere evaluated pathologically in 23 cases: 18 (78%) ofthe patients had metastatic foci. Residual MTC (de-tected by hypercalcitoninemia and/or general symp-toms 2 months after the initial surgery) was found in27 of the cases. In the remaining 9 patients, the firstsign of recurrent MTC (detected also by hypercalci-toninemia and/or general symptoms) was observedafter 32 months on average (range, 6 –103 months).Thirty-four patients had an elevated basal calcitoninlevel (mean, 1048.7 pg/mL; range, 10.1–5827 pg/mL),with (20 patients) or without (14 patients) an in-creased CEA level, and 9 had general symptoms (diar-rhea and/or flushing). In two patients, diarrhea and/orflushing were the only signs of recurrence.

The whole spectrum of conventional imaging mo-dalities before the initiation of the current study re-vealed a few enlarged ($ 1 cm) cervical, mediastinal,supraclavicular, and axillary LNs (16 patients; meannumber per patient, 2.5; range, 1–9), 3 cases of multi-ple small pulmonary metastases (detected by CT), 1case of solitary (identified by bone scan, MRI and SRS)and 1 case of multiple bony metastasis (diagnosed bybone scan and MRI), 1 case with breast metastasis(detected by mammography and surgical biopsy), and

5 cases with nondetermined hepatic lesions (1–5 foci,detected by CT, MRI, and US). 131I-meta-iodobenzyl-guanidine examinations revealed only 1 pathologicfocus (LN) among the total patient population.

18F-deoxyglucose PET demonstrated metastaticLNs in 34 (94%) patients (mean number per patient,8.9; range, 1–22), mainly in the cervical and mediasti-nal regions. Aspiration cytology or surgical interven-tion was performed in six patients after PET scancompletion, and the subsequent cytology/pathologyproved the tumorous involvement of the FDG-accu-mulating LNs in all six cases. Apart from two multiplepulmonary metastatic cases, all previously known me-tastases were revealed by FDG PET (not shown inTable 1), and viable tumor tissue indicating enhancedFDG accumulation also was detected in the liver (sixpatients with one to three lesions; Table 1).

The dynamic CT and MRI liver examinations re-sulted in a likewise modest number of pathologic he-patic foci. In 20 cases, no such foci were found. Mul-tiple foci were observed in four cases, but only in twocases did these imaging techniques provide sufficientinformation to diagnose multiple hypervascular he-patic metastases. In the remaining two cases, the le-sions also could be interpreted as foci of obscureorigin or cystic lesions. In 12 cases, 1 to 6 hepatic fociwere detected, but only in 5 cases did the investigatorsmake a firm statement of the metastatic origin of thefoci.

Analysis of the hepatic angiography images re-vealed hypervascular metastatic foci in 32 cases (89%),with a mean of 23.5 (range, 5–169) foci per patient;they were typically small (# 10 mm), with only 6 foci(among the total study population) measuring . 10mm. Liver images for a representative case (Patient 16)are shown in Figure 1. Hepatic angiography, revealinga high number (71) of hypervascular lesions of 3– 8mm in size, affords a strong contrast to dynamic CTand MRI in the different phases, as the latter togetherdid not indicate even a single lesion.

Three (Patients 26, 29, and 30) of the 4 patientsnot furnishing angiographic evidence of liver metas-tasis displayed a basal calcitonin level in the range18.7– 44.8 pg/mL. The patient follow-up has revealed astable clinical state (identical to that characterized bythe data in Table 1) in each of these 3 cases for the past2 years. Mediastinal and left cervical LN dissectionwas performed in Patient 5 (the fourth with negativehepatic angiography) after angiography. As a result ofthe surgical intervention, the calcitonin level was re-duced from 3556 to 874 pg/mL, and no indication ofpathologic LNs or any other metastasis was found onCT and MRI investigation during the systematic fol-low-up in the 2 years that has elapsed.


TABL

E1

Path

olog

icD

ata,

Mar

ker

Leve

lsan

dRe

sults

ofIm

agin

gof

Patie

nts

with

Resi

dual

orRe

curr

ing

Med

ulla

ryTh

yroi

dCa

rcin

oma

Patie

ntno

.

Initi

alsta

ging

Sign

ofre

curr

ence

No.o

fdia

gnos

edm

etas

tase

sby

conv

entio

nali

mag

ing

Unde

term

ined

hepa

ticfo

ci(n

)

Lym

phno

dem

etas

tase

s(b

yPET

)

Hepa

ticm

ets.

(by

PET)

(n)

Dete

cted

hepa

ticfo

ci

Dyna

mic

n/or

igin

b

Angi

ogra

phy

(hyp

erva

scul

arm

etas

tase

s)pT (c

m)

Path

olog

icly

mph

node

invo

lvem

enta

Basa

lca

lcito

nin

pg/m

LCE

A(n

g/m

L)Fl

ush/

diar

rhea

Lym

phno

dePu

lmon

ary

Bone

Othe

rsn

Loca

tion

CTM

RIn

Size

rang

e

11.4

158

277.9

1—

——

——

6l-r

C,M

—1/

cy1/

cy11

3–7

26.0

140

8526

0.01

3m

pl—

—2

12l-r

C,M

,—

mpl

/m-q

mpl

/m-q

169

2–10

36.0

139

7441

.92

4—

——

—5

l-cC,

M—

——

313–

54

4.01

3802

130.5

2—

——

—3

9r-c

C,M

—1/

q1q

273–

355

4.0n.

p.35

5669

4.51

9—

——

—14

lC,l

S,M

——

—0

—6

4.51

1844

123.8

26

——

——

21l-r

-cC,

M3

——

93–

57

6.0n.

p.16

3115

7.92

1—

mpl

lbre

ast

—4

M3

mpl

/mm

pl/m

333–

108

2.01

1627

6.51

2m

pl—

——

19l-r

-cC,

rS,M

—2/

q2/

q5

3–5

92.0

112

9035

.72

1—

——

111

l-cC,

M—

3/q

3/q

144–

1010

5.02

(11)

1146

6.32

2—

——

—6

r-cC,

rS,M

—1/

cy1/

cy41

2–5

114.2

2(4

)10

4119

.22

1—

——

113

l-cC,

M—

1/m

1/m

203–

1012

2.01

1007

6.12

——

——

—4

rC,M

——

—15

3–5

132.5

110

2213

.32

1—

——

—11

l-r-c

C,M

—2/

q2/

q14

3–8

144.5

169

024

.62

——

——

—5

cC,M

—2/

m2/

m10

2–5

151.0

152

67.9

21

——

——

22l-r

-cC,

l-rS,

l-rA,

M2

——

113–

516

4.52

(15)

373

250.5

1—

——

——

16l-r

-cC,

l-rS,

M—

——

713–

817

2.0n.

p.36

36.4

2—

——

——

1rC

—1/

cy1/

cy9

318

6.0n.

p.29

89.4

2—

——

—5

3r-c

C—

6/m

-q6/

m-q

85–

3019

2.01

288

8.22

1—

——

—14

l-r-c

C,rS

,M1

——

93–

1020

7.02

(18)

279

2.32

——

——

—2

l-rC

——

—8

3–5

212.0

126

110

.12

2—

——

—20

l-r-c

C,lS

,M—

——

53–

422

4.51

199

1.22

1—

——

—9

l-r-c

C—

——

63–

423

2.01

184

5.01

3—

——

—10

l-rC,

M3

——

53–

424

1.6n.

p.90

.83.3

1—

——

——

1rC

—1/

m1/

m25

3–15

252.0

n.p.

560.1

12

——

——

6l-r

C,M

—m

pl/c

ym

pl/c

y12

3–5

266.0

144

.81.4

2—

——

——

7cC

,M—

——

0—

273.0

n.p.

40.2

0.32

——

——

—0

——

mpl

/qm

pl/q

119

2–3

281.4

2(1

2)24

1.92

——

——

2lC

——

—17

2–5

293.0

119

1.32

——

——

—5

cC,M

——

n.p.

0—

301.0

n.p.

18.7

0.62

——

solit

ary

——

10l-c

C,M

3—

—0

—31

2.51

17.9

1.22

——

——

—7

l-cC

——

—5

332

4.4n.

p.10

.21.2

2—

——

——

3l-r

C—

——

64–

533

1.6n.

p.10

.11.2

1—

——

——

7l-r

C,M

——

—9

3–5

344.6

n.p.

10.1

3.42

——

——

—7

l-rC,

M—

——

113–

535

2.0n.

p.0.1

0.71

—m

pl—

——

11l-r

-cC,

M—

——

62–

336

1.8n.

p.3.9

0.21

——

——

—0

——

1/m

1/m

113–

8

A:ax

illar

y;c:

cent

ral;

C:ce

rvica

l;cy

:cys

t;l:

left;

m:m

etas

tasis

;M:m

edia

stina

l;m

pl:m

ultip

le;n.

p.:n

otpe

rform

ed;r

:rig

ht;S

:sup

racla

vicul

ar;q

:que

stion

able.

aTh

enu

mbe

rofn

egat

ivelym

phno

desi

sgive

nin

pare

nthe

ses.

bTh

ehe

patic

lesio

nm

aybe

met

asta

sis,c

yst,

orqu

estio

nabl

e.


In Patients 2 and 16, with 169 and 71 angiographi-cally detected lesions, respectively, laparoscopy dem-onstrated only 3 and 5 small (, 2 mm) bright whitishfoci, respectively, on the hepatic surface, which werelater proved by immunophenotyping to be MTC me-tastases. In Patients 25, 33, and 36 with 12, 9, and 11angiographically detected liver metastases, respec-tively, no lesions were observed on the liver surface,and the biopsy proved to be negative for tumor.

DISCUSSIONMany publications report that the plasma calcitoninconcentration roughly reflects the tumor burden inMTC patients.28 –31 However, the frequent discrepancybetween the relatively low tumor mass detected andthe high tumor marker level in many MTC patientsalso has long been known.27,28,32,33 The explanation ofthis mismatch may lie in the heterogenity of the tumorcell population and in the decrease in tumor markerexpression during dedifferentiation of MTC,27,32 but italso may be explained by undiscovered small metas-tases.

Medullary thyroid carcinoma belongs in the groupof neuroendocrine tumors. It is well documented thatthe natural history of these tumors includes a pro-nounced early lymphatic spreading and hepatic dis-semination of typical hypervascular lesions suppliedby the branches of the a. hepatica propria.34,35 Thus, inMTC too, hepatic involvement is expected to be ahighly probable phenomenon, though surprisinglythis never has been proved.

In an attempt to explore all tumor-infiltrated tis-sue in our basically hypercalcitoninemic MTC pa-tients, we conducted a regular search for metastases.Conventional spiral CT, MRI, US, and MIBG, SR andbone scintigraphy were applied first to detect meta-static LNs, and hepatic, pulmonary, and bony involve-ment, the most frequent metastases in MTC. As anoteworthy result, these studies revealed only a rela-tively few foci of metastatic disease (Table 1). Theresults of these imaging modalities are reliable forpulmonary (CT) and bony (MRI and SRS) metastasesbut questionable in cases of LN metastases because inanatomic imaging (CT, MRI, and US) size is the onlycriterion, and metabolic imaging using a catechol-amine precursor (MIBG) or SR specific ligand has rel-atively low specificity. Very few publications report onliver metastases detected with any of these modalities,and we did not obtain convincing results on the livereither (see Table 1). The failure to localize both LN andhepatic metastases, potentially responsible for highmarker levels might be related to the limited sensitiv-ity of the applied techniques to detect tumorous le-

sions in these regions;10,36 thus, other diagnosticmethods were needed.

Scintigraphic imaging methods such as 201Tl-chlo-ride,37 99mTc(V)-dimercaptosuccinic acid,37–39 99mTc-sestamibi,37,40 131I or 123I-labeled MIBG,41 and labeledsomatostatin analogs38,39,42– 48 dramatically underesti-mate the integrated tumor mass in MTC, because theirsensitivity is rather low, and the specificity of some ofthe tracers in detecting metastases is also limited.Although labeled anticalcitonin49 and labeled anticar-cinoembryonic43,50,51 antigen antibodies are highlyspecific radiopharmaceuticals, and their applicationmay ensure high sensitivity, they are seldom usedbecause of their inherent immunogenicity and thedifficulties in radiolabeling. We tested only MIBG fromamong the aforementioned tracers and obtained al-most entirely negative results: among the 36 patientswith signs of recurrent MTC, only a single pathologicfocus was observed. Thus, we concluded that MIBGscintigraphy is of very low value in the searching fortumorous foci in patients with recurring MTC.

Positron emission tomography, with its outstand-ing viable tumor detection sensitivity, was chosen tomaximize tumorous involvement localization. Therehave been few PET studies in MTC;52–54 these docu-ment the marked sensitivity of this method in localiz-ing infiltrated LNs, the results in detecting bony andpulmonary metastases being not so good, though stillacceptable. In contrast, PET does not reliably localizeliver involvement, a finding explained by the highaccumulation of the PET tracer (labeled FDG or me-thionine) in the liver. Our whole-body FDG PET inves-tigations of the 36 subjects revealed lymphatic dissem-ination in 34 patients (94%). Besides LN involvement,six patients were identified with liver metastasis, onepatient with pulmonary metastasis (two cases werefalse-negative), two patients with bony metastasis,and one patient with breast metastasis. The high in-cidence of LN metastases in our study may be relatedto the advanced stage of the disease because of theselection criteria (patients with recurrences), the largeaverage dimension of the primary tumors (3.3 cm,close to pT3 tumor size), and a long mean follow-up(45 months from the detection of residual or recurrenttumorous signs). The dimension of the FDG-accumu-lating LNs was less than 1 cm in most cases, a findingconsistent with the rare LN involvement identified byusing CT and MRI. The results of our PET studies withthe detected involvements of many LNs and a fewother metastases still did not lead to matching be-tween the plasma tumor marker level/general symp-toms and the total tumor mass. To resolve this issue,additional possible metastatic sites had to be identi-fied by procedures superior to those already applied.


FIGURE 1. Liver images of Patient 16. (a) Unenhanced computed tomography (CT), (b) contrast-enhanced arterial phase CT; (c) contrast-enhanced portal phase

CT; (d) T2-weighted magnetic resonance imaging (MRI), (e) T1-weighted MRI, (f) contrast-enhanced, T1-weighted arterial phase MRI; (g) contrast-enhanced,

T1-weighted portal phase MRI; (h) hepatic angiography, arterial phase; (i) hepatic angiography, parenchymal phase. Hepatic angiography, revealing a high number

(71) of hypervascular lesions of 3–8 mm in size, affords a strong contrast to dynamic CT and MRI in the different phases, because the latter together did not indicate

even a single lesion. Postangiographic laparoscopy revealed 5 small (, 2 mm) bright whitish foci on the hepatic surface, which were later proved by

immunophenotyping to be medullary thyroid carcinoma metastases.


Because to our knowledge there are no systematicstudies of this issue in MTC, we hypothesized that, aswith other types of neuroendocrine tumors, hypervas-cular, small hepatic metastases are characteristic forMTC also. The tumor-detecting sensitivity of all radio-nuclide methods is rather poor for such metastases,and anatomic imaging with simple conventional pro-tocols is not high enough either. Accordingly, wechose additional diagnostic procedures. Previousstudies have revealed that dynamic CT and MRI meth-ods are more effective for the demonstration of hyper-vascularized liver lesions than static measuring proto-cols. Of the dynamic CT examinations, triple-phasehelical examinations (unenhanced pictures, and thenarterial and portal phase studies after a bolus injectionof contrast medium) proved to be the most sensitiveprotocols,22,23,26,55–58 though the sensitivity was stillnot high enough. The highest rate of detection can beachieved when the information is available from allthree phases of the investigation, because sufficientlesion characterization requires all three sets ofimages.

Advances in MRI24,25,58 (breath-hold sequences,application of extracellular contrast agents) have re-sulted in its increasing use for abdominal imaging. Forthe detection of hepatic metastases, T1- and T2-weighted breath-hold sequences with multiphase gad-olinium enhanced dynamic study of the entire liver isthe recommended imaging protocol. Lesion charac-terization is possible by using the information of dif-ferent sequences and the contrast-enhancement pat-tern during arterial, portal, and equilibrum phases.59

Comparative studies showed that this MRI method ismore sensitive and specific than dynamic helical CTexamination.60 – 62

For technical reasons, the CT and MRI methodsmentioned above are inherently of low sensitivity forthe detection of lesions less than 5 mm in dimension.To cover the entire liver, we performed the CT exam-inations with 6 –10 mm collimation, and the MRI se-quences had a slice thickness of 5–10 mm. Partialvolume effects prevent the detection and more oftenthe characterization of lesions measuring less than 5mm. The recent development of multidetector-row CTscanners63 and tissue specific MRI contrast agents25,58

appears to be of promise for the detection of smallhypervascular metastases.

The data in Table 1 clearly show the increasedliver metastasis-detecting sensitivity of dynamic CTand MRI investigations relative to that of the moreconventional CT and MRI protocols. These dynamicmethods have comparable sensitivities: metastaseswere detected either by both or by neither of them.Nevertheless, the lack of evidence of liver metastasesFIGURE 1. (continued)


by these procedures in more than two-thirds of thecases did not resolve the contradiction between thesmall tumor mass and the high plasma level of tumormarkers or the presence of general symptoms in ourpatients.

In consequence of the low effectivity of all theavailable noninvasive methods in the diagnosis ofsmall hypervascular hepatic metastases, we attemptedto use invasive procedures, as well. Selective venouscatheterization (SVC)13,16-19,64 – 66 has long been themethod of choice for the approximate localization oftumor metastases in MTC patients with persistent orrecurrent hypercalcitoninemia and no clear-cut evi-dence of tumorous masses when conventional diag-nostic imaging methods are used. The calcitonin mea-surements may be performed either in thenonstimulated state,13,64 or before and after stimula-tion tests (applying pentagastrin or calcium).16 –19,66 Ifthe calcitonin concentration in a given vein is in thepathologic range17–19,66 or the calcitonin gradient ex-ceeds 1.213,16 or more strictly 1.564 relative to the meanperipheral vein calcitonin level, the SVC test is re-garded as positive. Although SVC does not belongamong the morphologic methods and is not capableof detecting the tumor mass itself, it can confirm cal-citonin secretion by even a very tiny focus. However,the results of SVC often are not consistent with thoseof other methods.13,16,18,43,45,64 The low effectivity ofthe SVC data-based secondary surgical interventionson MTC patients with persistent or recurrent hyper-calcitoninemia2,8,11–20 also points to the limited appli-cability of this localizing method in MTC.

A small fraction of the SVC studies addressed theexploration of hepatic involvement and resulted infew positive hepatic gradients: 4/19,13 4/1764 and2/14.16 Only a proportion of the cases with pathologicSVC results were proved to involve hepatic metastasesby histology, angiography, or conventional diagnosticimaging before the completion of the appropriatestudies.13,16,64 The only systematic investigation onthis issue concerned 13 surgically treated MTC pa-tients.66 Gautvik et al. reported that the increase in thecalcitonin level in the hepatic vein was more pro-nounced than that in the peripheral veins and fol-lowed the stimulation more quickly than the latter inall cases.66 Liver metastases, however, were localizedby these authors by means of imaging methods (CTand scintigraphy) in only four patients. Nevertheless,these result strongly suggested that MTC frequentlyspreads to the liver and in addition indicated the dif-ficulty of localizing metastatic masses by conventionalimaging methods.

Several groups have applied diagnostic laparos-copy or open laparotomy with biopsy to search for

liver metastases.17,67– 69 The typical appearance of he-patic MTC lesions is multiple, tiny (1–5 mm), brightwhitish foci. Moley et al. and Tung et al. advocatelaparoscopic liver examination in all patients with ahigh peak peripheral stimulated plasma calcitoninlevel to disclose the presence of hepatic metastasesbefore extensive surgical intervention.17,68 Tung re-ported a low incidence (8 of 41; 19.5%) of laparoscopy-proven hepatic metastases68 in his systematic study of41 patients with persistent or recurrent hypercalci-toninemia. More interestingly, SVC revealed 10 pa-tients within this study population with an increasedhepatic vein calcitonin level, only 2 of whom (20%)had laparoscopy (i.e., pathologically) proven hepaticmetastases. This latter finding clearly demonstratesthat SVC does not identify all liver metastases (be-cause six patients with pathologically proven metas-tases were SVC negative), and that although laparos-copy is a particularly sensitive tool for theidentification of pathologically unsuspected lesionsalong the liver surface, it fails in cases of nonsurface-sitting hepatic metastases.

Liver angiography is a very effective method forthe localization of occult hepatic metastases in MTC,because it is capable of detecting very small (severalmillimeters), nonsurface-sitting lesions of the liv-er.65,69 It unequivocally surpasses laparoscopy with itsability to reveal cryptic metastases within the liverparenchyma, but it does not provide pathologic ma-terial. In addition, the characteristic angiographic ap-pearance of the metastatic foci and their multiplicitystrongly support the diagnosis. No systematic studyhas been published on angiography as a diagnosticpossibility in MTC, and we therefore decided to scru-tinize the value of this invasive diagnostic imagingprocedure in the detection of hepatic metastases. Inall but 4 cases (89%), our liver angiography examina-tions confirmed multiple small hypervascular, hepaticmetastases that were practically undetectable withother imaging modalities (see Table 1 and Fig. 1).

The finding that liver angiography did not give apositive result in Patients 5, 26, 29, and 30 might beexplained by the limited sensitivity of this imagingmodality in detecting small lesions. Liver metastasesmeasuring 1–2 mm may be below the detection limitof the method. This explanation may be supported bythe relatively low elevation in the plasma calcitoninconcentration in Patients 26, 29, and 30, suggesting anot too high tumor burden. The sustained very highcalcitonin level in Patient 5, which persisted even afterthe second surgical intervention, with negative post-operative diagnostic imaging studies, may have re-sulted from an extremely high number of subthresh-old sized hepatic lesions.


Our observations demonstrate that liver metasta-ses are common in MTC, and their almost regular butinconspicuous occurrence offers an obvious explana-tion for the calcitonin level—tumor mass mismatch.The low rate of postsurgical cure after secondary sur-gical (mainly LN) interventions2,8,11–20 also becomesquite understandable in the knowledge of the ex-tremely frequent liver metastases. The high incidenceof liver involvement detected by hepatic angiographyindicates that the relatively low prevalence of distantmetastases reported by numerous groups using CT,MRI, US, and scintigraphic methods is very probably amisconclusion in both the initial and advanced stagesof the disease.

The similar incidences of lymphatic (94%) andhepatic (89%) dissemination suggest a possible paral-lel development of hepatic and lymphatic spreading.As the incidence of the pathologic involvement of LNsrelates to the size of the primary tumor,2 hepatic dis-semination may be predicted by the latter. The pre-sented data suggest that hepatic metastases developwith high probability if the primary tumor dimensionexceeds 1 cm (although the smallest reported primarytumor accompanied by hepatic metastases was 1mm68). This compares favorably with reports on thefrequent hepatic dissemination in other neuroendo-cine tumors (carcinoid, gastrinoma, and islet cell tu-mor) when the primary tumor is larger than 1 cm.69,70

Hepatic metastatization occurred in our patients evenwith normal basal plasma marker levels. Liver angiog-raphy therefore should be recommended for correctstaging in all MTC cases with a primary tumor exceed-ing 1 cm and/or pathologic involvement of LNs, irre-spective of the basal plasma tumor marker levels. Lo-calization of all metastatic foci is important during thefollow-up, because patients with widespread dissem-inated disease should be spared from unnecessarysurgical interventions and may be forwarded to sys-temic radionuclide (somatostatin analog or MIBG)therapy.

REFERENCES1. Beressi N, Campos JM, Beressi JP, Franc B, Niccoli-Sire P,

Conte-Devolx B, et al. Sporadic medullary microcarcinomaof the thyroid: a retrospective analysis of eighty cases. Thy-roid 1998;8:1039 – 44.

2. Gimm O, Ukkat J, Dralle H. Determinative factors of bio-chemical cure after primary and reoperative surgery forsporadic medullary thyroid carcinoma. World J Surg 1998;22:562– 8.

3. Høie J, Jørgensen OG, Stenwig AE, Langmark F. Medullarythyroid cancer in Norway. Acta Chir Scan 1988;154:339 – 43.

4. Moley JF, DeBenedetti MK. Patterns of nodal metastases inpalpable medullary thyroid carcinoma. Recommendationsfor extent of node dissection. Ann Surg 1999;229:880 – 8.

5. Schroder S, Bocker W, Baisch H, Burk CG, Arps H, Meiners

I, et al. Prognostic factors in medullary thyroid carcinomas.Survival in relation to age, sex, stage, histology, immunocy-tochemistry, and DNA content. Cancer 1988;61:806 –16.

6. Scopsi L, Sampietro G, Boracchi P, Del Bo R, Gullo M,Placucci M, et al. Multivariate analysis of prognostic factorsin sporadic medullary carcinoma of the thyroid. A retro-spective study of 109 consecutive patients. Cancer 1996;78:2173– 83.

7. Simpson WJ, Palmer JA, Rosen IB, Mustard RA. Managementof medullary carcinoma of the thyroid. Am J Surg 1982;144:420 –2.

8. Dralle H, Damm I, Scheumann GFW, Kotzerke J, Kupsch E,Geerlings H, et al. Compartment-oriented microdissectionof regional lymph nodes in medullary thyroid carcinoma.Surg Today 1994;24:112–21.

9. Dralle H, Scheumann GFW, Proye C, Bacourt F, Frilling A,Limbert F, et al. The value of lymph node dissection inhereditary medullary thyroid carcinoma: a retrospective,European, multicentre study. J Int Med 1995;238:357– 61.

10. Fraker DL, Skarulis M, LiVolsi V. Thyroid tumors. In: DeVitaV Jr., Hellman S, Rosenberg SA, editors. Cancer: principlesand practice of oncology. 5th ed. Philadelphia, New York:Lippincott-Raven, 1997:1629 –52.

11. Kallinowski F, Buhr HJ, Meybier H, Eberhardt M, Herfarth C.Medullary carcinoma of the thyroid—therapeutic strategyderived from fifteen years of experience. Surgery 1993;114:491– 6.

12. Buhr HJ, Kallinowski F, Raue F, Frank-Raue K, Herfarth C.Microsurgical neck dissection for metastasizing medullarythyroid carcinoma. Eur J Surg Oncol 1995;21:195–7.

13. Abdelmoumene N, Schlumberger M, Gardet P, Roche A,Travagli JP, Francese C, et al. Selective venous samplingcatheterisation for localisation of persisting medullary thy-roid carcinoma. Br J Cancer 1994;69:1141– 4.

14. Block MA, Jackson CE, Tashjian AH Jr. Management of occultmedullary thyroid carcinoma. Arch Surg 1978;113:368–72.

15. Ellenhorn JDI, Shah JP, Brennan MF. Impact of therapeuticregional lymph node dissection for medullary carcinoma ofthe thyroid gland. Surgery 1993;114:1078 – 82.

16. Frank-Raue K, Raue F, Buhr HJ, Baldauf G, Lorenz D, ZieglerR. Localization of occult persisting medullary thyroid carci-noma before microsurgical reoperation: high sensitivity ofselective venous catheterization. Thyroid 1992;2:113–7.

17. Moley JF, Dilley WG, DeBenedett MK. Improved result ofcervical reoperation for medullary thyroid carcinoma. AnnSurg 1997;225:734 – 43.

18. Norton JA, Doppman JL, Brennan MF. Localization andresection of clinically inapparent medullary carcinoma ofthe thyroid. Surgery 1980;87:616 –22.

19. Tisell L-E, Hansson G, Jansson S, Salander H. Reoperation inthe treatment of asymptomatic metastasizing medullarythyroid carcinoma. Surgery 1986;99:60 – 6.

20. van Heerden JA, Grant CS, Gharib H, Hay ID, Ilstrup DM.Long-term course of patients with persistent hypercalci-toninemia after apparent curative primary surgery for med-ullary thyroid carcinoma. Ann Surg 1990;212:395– 401.

21. Imran MB, Kubota K, Yamada, S, Fukuda H, Yamada K,Fujiwara T, et al. Lesion-to-background ratio in nonattenu-ation-corrected whole-body FDG PET images. J Nucl Med1998;39:1219 –23.

22. Oliver JH III, Baron RL, Federle MP, Jones BC, Sheng R.Hypervascular liver metastases: do unenhanced and hepaticarterial phase CT images affect tumor detection? Radiology1997;205:709 –15.


23. Paulson EK, McDermott VG, Keogan MT, DeLong DM, Fred-erick MG, Nelson RC. Carcinoid metastases to the liver: roleof triple-phase helical CT. Radiology 1998;206:143–50.

24. Soyer P, de Givry SC, Gueye C, Lenormand S, Somveille E,Scherrer A. Detection of focal hepatic lesions with MR im-aging: prospective comparison of T2-weighted fast spin-echo with and without fat suppression, T2-weighted breath-hold fast spin-echo, and gadolinium chelate-enhanced 3Dgradient-recalled imaging. Am J Roentgenol 1996;166:1115–21.

25. van Beers BE, Gallez B, Pringot J. Contrast-enhanced MRimaging of the liver. Radiology 1997;203:297–306.

26. van Leeuwen MS, Noordzij J, Feldberg MAM, HennipmanAH, Doornewaard H. Focal liver lesions: characterizationwith triphasic spiral CT. Radiology 1996;201:327–36.

27. Busnardo B, Girelli ME, Simioni N, Nacamulli D, Busetto E.Nonparallel patterns of calcitonin and carcinoembryonicantigen levels in the follow-up of medullary thyroid carci-noma. Cancer 1984;53:278 – 85.

28. Block MA, Jackson CE, Tashjian AH Jr. Medullary thyroidcarcinoma detected by serum calcitonin assay. Arch Surg1972;104:579 – 86.

29. Tisell LE, Dilley WG, Wells SA Jr. Progression of postopera-tive residual medullary thyroid carcinoma as monitored byplasma calcitonin levels. Surgery 1996;119:34 –9.

30. Wells SA Jr., Baylin SB, Leight GS, Dale JK, Dilley WG, Farn-don JR. The importance of early diagnosis in patients withhereditary medullary thyroid carcinoma. Ann Surg 1982;195:595–9.

31. Wells SA Jr., Baylin SB, Gann DS, Farrell RE, Dilley WG,Preissig SH, et al. Medullary thyroid carcinoma: relationshipof method of diagnosis to pathologic staging. Ann Surg1978;188:377– 83.

32. Lippman SM, Mendelsohn G, Trump DL, Wells SA Jr., BaylinSB. The prognostic and biological significance of cellularheterogeneity in medullary thyroid carcinoma: a study ofcalcitonin, L-dopa decarboxylase, and histaminase. J ClinEndocrinol Metab 1982;54:233– 40.

33. Trump DL, Mendelsohn G, Baylin SB. Discordance betweenplasma calcitonin and tumor-cell mass in medullary thyroidcarcinoma. N Eng J Med 1979;301:253–5.

34. Fraker DL, Jensen RT. Pancreatic endocrine tumors. In: De-Vita VT Jr., Hellman S, Rosenberg SA, editors. Cancer: prin-ciples and practice of oncology. 5th ed. Philadelphia, NewYork: Lippincott-Raven, 1997:1678 –704.

35. Jensen RT, Norton JA. Carcinoid tumors and the carcinoidsyndrome. In: DeVita VT Jr., Hellman S, Rosenberg SA, ed-itors. Cancer: principles and practice of oncology. 5th ed.Philadelphia, New York: Lippincott-Raven, 1997:1704 –23.

36. Marsh DJ, Learoyd DL, Robinson BG. Medullary thyroidcarcinoma: recent advances and management update. Thy-roid 1995;5:407–24.

37. Ugur O, Kostakoglu L, Guler N, Caner B, Uysal U, Elahi N, etal. Comparison of 99mTc(V)-DMSA, 201TI and 99mTc-MIBIimaging in the follow-up of patients with medullary carci-noma of the thyroid. Eur J Nucl Med 1996;23:1367–71.

38. Adams S, Baum RP, Hertel A, Schumm-Draeger PM, UsadelK-H, Hor G. Comparison of metabolic and receptor imagingin recurrent medullary thyroid carcinoma with histopatho-logical findings. Eur J Nucl Med 1998;25:1277– 83.

39. Kurtaran A, Scheuba C, Kaserer K, Schima W, Czerny C,Angelberger P, et al. Indium-111-DTPA-D-Phe-1-octreotideand technetium-99m-(V)-dimercaptosuccinic acid scanning

in the preoperative staging of medullary thyroid carcinoma.J Nucl Med 1998;39:1907–9.

40. Learoyd DL, Roach PJ, Briggs GM, Delbridge LW, Wilms-hurst EG, Robinson BG. Technetium-99m-sestamibi scan-ning in recurrent medullary thyroid carcinoma. J Nucl Med1997;38:227–30.

41. Troncone L, Rufini V, Montemaggi P, Danza FM, Lasorella A,Mastrangelo R. The diagnostic and therapeutic utility ofradioiodinated metaiodobenzylguanidine (MIBG). 5 years ofexperience. Eur J Nucl Med 1990;16:325–35.

42. Baudin E, Lumbroso J, Schlumberger M, Leclere J, Giam-marile F, Gardet P, et al. Comparison of octreotide scintig-raphy and conventional imaging in medullary thyroid car-cinoma. J Nucl Med 1996;37:912– 6.

43. Behr TM, Gratz S, Markus PM, Dunn RM, Hufner M,Schauer A, et al. Anti-carcinoembryonic antigen antibodiesversus somatostatin analogs in the detection of metastaticmedullary thyroid carcinoma. Are carcinoembryonic anti-gen and somatostatin receptor expression prognostic fac-tors? Cancer 1997;80:2436 –57.

44. Berna L, Chico A, Matias-Guiu X, Mato E, Catafau A, AlonsoC, et al. Use of somatostatin analogue scintigraphy in thelocalization of recurrent medullary thyroid carcinoma. EurJ Nucl Med 1998;25:1482– 8.

45. Dorr U, Sautter-Bihl M-L, Bihl H. The contribution of soma-tostatin receptor scintigraphy to the diagnosis of recurrentmedullary carcinoma of the thyroid. Semin Oncol 1994;21:42–5.

46. Frank-Raue K, Bihl H, Dorr U, Buhr H, Ziegler R, Raue F.Somatostatin receptor imaging in persistent medullary thy-roid carcinoma. Clin Endocrinol 1995;42:31–7.

47. Krenning EP, Kwekkeboom DJ, Bakker WH, Breeman WAP,Kooij PPM, Oei HY, et al. Somatostatin receptor scintigraphywith [111In-DTPA-D-Phe1]- and [123I-Tyr3]-octreotide: theRotterdam experience with more than 1000 patients. EurJ Nucl Med 1993;20:716 –31.

48. Kwekkeboom DJ, Reubi JC, Lamberts SWJ, Bruining HA,Mulder AH, Oei HY, et al. In vivo somatostatin receptorimaging in medullary thyroid carcinoma. J Clin EndocrinolMetab 1993;76:1413–7.

49. Manil L, Boudet F, Motte P, Gardet P, Saccavini J-C, Lum-broso J-D, et al. Positive anticalcitonin immunoscintigraphyin patients with medullary thyroid carcinoma. Cancer Res1989;49:5480 –5.

50. Barbet J, Peltier P, Bardet S, Vuillez JP, Bachelot I, Denet S,et al. Radioimmunodetection of medullary thyroid carci-noma using indium-111 bivalent hapten and anti-CEA Xanti-DTPA-indium bispecific antibody. J Nucl Med 1998;39:1172– 8.

51. Peltier P, Curtet C, Chatal J-F, Le Doussal J-M, Daniel G,Aillet G, et al. Radioimmunodetection of medullary thyroidcancer using a bispecific anti-CEA/anti-indium-DTPA anti-body and an indium-111-Labeled DTPA dimer. J Nucl Med1993;34:1267–73.

52. Brandt-Mainz K, Muller SP, Gorges R, Saller B, Bockisch A.The value of fluorine-18 fluorodeoxyglucose PET in patientswith medullary thyroid cancer. Eur J Nucl Med 2000;27:490 – 6.

53. Gasparoni P, Rubello D, Ferlin G. Potential role of fluorine-18-deoxyglucose (FDG) positron emission tomography(PET) in the staging of primitive and recurrent medullarythyroid carcinoma. J Endocrinol Invest 1997;20:527–30.


54. Musholt TJ, Musholt PB, Dehdashti F, Moley JF. Evaluationof fluorodeoxyglucose-positron emission tomographicscanning and its association with glucose transporter ex-pression in medullary thyroid carcinoma and pheochromo-cytoma: a clinical and molecular study. Surgery 1997;122:1049 – 60.

55. Choi BI, Lee HJ, Han JK, Choi DS, Seo JB, Han MC. Detectionof hypervascular nodular hepatocellular carcinomas: valueof triphasic helical CT compared with iodized-oil CT. Am JRoentgenol 1997;168:219 –24.

56. Huenerbein R, Reuter P, Kuhn F-P. Value of noncon-trast phase in detecting hypervascular metastases withtriphasic helical CT? [letter, comment]. Radiology 1998;209:585– 6.

57. Miller FH, Butler RS, Hoff FL, Fitzgerald SW, Nemcek AA Jr.,Gore RM. Using triphasic helical CT to detect focal hepaticlesions in patients with neoplasms. Am J Roentgenol 1998;171:643–9.

58. Sica GT, Ji H, Ros PR. CT and MR imaging of hepatic me-tastases. Am J Roentgenol 2000;174:691– 8.

59. Soyer P, Gueye C, Somveille E, Laissy J-P, Scherrer A. MRdiagnosis of hepatic metastases from neuroendocrine tu-mors versus hemangiomas: relative merits of dynamic gad-olinium chelate-enhanced gradient-recalled echo and un-enhanced spin-echo images. Am J Roentgenol 1995;165:1407–13.

60. Larson RE, Semelka RC, Bagley AS, Molina PL, Brown ED,Lee JK. Hypervascular malignant liver lesions: comparisonof various MR imaging pulse sequences and dynamic CT.Radiology 1994;192:393–9.

61. Semelka RC, Shoenut JP, Kroeker MA, Greenberg HM, SimmFC, Minuk GY, et al. Focal liver disease: comparison ofdynamic contrast-enhanced CT and T2-weighted fat-sup-

pressed, FLASH, and dynamic gadolinium-enhanced MRimaging at 1.5 T. Radiology 1992;184:687–94.

62. Semelka RC, Worawattanakul S, Kelekis NL, John G, Woos-ley JT, Graham M, et al. Liver lesion detection, characteriza-tion, and effect on patient management: comparison ofsingle-phase spiral CT and current MR techniques. J MagnReson Imaging 1997;7:1040 –7.

63. Foley WD, Mallisee TA, Hohenwalter MD, Wilson CR, QuirozFA, Taylor AJ. Multiphase hepatic CT with a multirow de-tector CT scanner. Am J Roentgenol 2000;175:679 – 85.

64. Mrad MDB, Gardet P, Roche A, Rougier P, Calmettes C,Motte P, et al. Value of venous catheterization and calcito-nin studies in the treatment and management of clinicallyinapparent medullary thyroid carcinoma. Cancer 1989;63:133– 8.

65. Sizemore GW. Medullary carcinoma of the thyroid gland.Semin Oncol 1987;14:306 –14.

66. Gautvik KM, Talle K, Hager B, Jørgensen OG, Aas M. Earlyliver metastases in patients with medullary carcinoma of thethyroid gland. Cancer 1989;63:175– 80.

67. Babineau TJ, Lewis WD, Jenkins RL, Bleday R, Steele GJ Jr.,Forse RA. Role of staging laparoscopy in the treatment ofhepatic malignancy. Am J Surg 1994;167:151–5.

68. Tung WS, Vesely TM, Moley JF. Laparoscopic detection ofhepatic metastases in patients with residual or recurrentmedullary thyroid cancer. Surgery 1995;118:1024 –30.

69. Weber HC, Venzon DJ, Lin J-T, Fishbein VA, Orbuch M,Strader DB, et al. Determinants of metastatic rate and sur-vival in patients with Zollinger-Ellison syndrome (ZES): aprospective long-term study. Gastroenterology 1995;108:1637– 49.

70. Moertel CG. An Odyssey in the land of small tumors. J ClinOncol 1987;5:1503–22.


Documents

Angiography effectively supports the diagnosis of hepatic metastases in medullary thyroid carcinoma