DOI:10.2214/AJR.05.1529
AJR 2006; 187:571-574
© American Roentgen Ray Society
Percutaneous MR-Guided Radiofrequency Ablation of Recurrent Sacrococcygeal Chordomas
Volker Teichgräber1,2,
Christophe Aubé1,
Diethard Schmidt1,
Ekkehard Jehle3,
Claudius König1,
Claus D. Claussen1 and
Philippe L. Pereira1
1 Department of Radiology, University of Tübingen, Hoppe-Seyler-Strasse 3,
70276 Tübingen, Germany.
2 Present address: Department of Medical Oncology/Tumor Research, West German
Cancer Center, Essen, Germany.
3 Department of Surgery, University of Tübingen, Tübingen,
Germany.
Received August 30, 2005;
accepted after revision November 3, 2005.
Address correspondence to P. L. Pereira
(philippe.pereira{at}med.uni-tuebingen.de).
Abstract
OBJECTIVE. Radiofrequency ablation is emerging as a therapeutic
technique for the treatment of an increasing variety of tumors. Exact visual
guidance to the tumor and controlled delivery of energy is pivotal for
ablation success.
CONCLUSION. Introducing MRI as a guidance technique ideally uses
tumor-specific tissue characteristics, allows direct multiplanar
reconstruction for precise needle positioning, and permits real-time
monitoring and assessment of treatment-induced tissue signal changes to
increase the safety of the procedure.
Keywords: ablation • dynamic MRI • interventional radiology • MRI • percutaneous • radiofrequency
Introduction
Radiofrequency ablation is a minimally invasive technique used for the
treatment of liver malignancies. Reports of approaches using radiofrequency
ablation in the treatment of nonhepatic neoplasms have been published
[1,
2]. For percutaneous
radiofrequency ablations of liver tumors, sonography or CT is commonly used
for guidance. Recent reports have already shown the successful coagulation of
a chordoma using CT [3,
4]. We prefer to perform
radiofrequency ablation of chordomas under MRI guidance because of
tissue-specific characteristics, direct multiplanar reconstruction, and the
near real-time monitoring of tissue changes. The latter feature represents our
major argument in favor of MR guidance because precise ablation algorithms are
missing [5]. Because of their
high water content, chordomas may present with different electrical properties
compared with liver tumors, and algorithms for extrahepatic tumor ablation are
seldom reported in the literature
[6]. We describe the successful
ablation of two recurrent sacrococcygeal chordomas under MR guidance to show
the potential of MR-guided radiofrequency ablation.
Materials and Methods
A 66-year-old woman was diagnosed in 1998 with sacrococcygeal chordomas.
From 1998 to 2001, she underwent four surgical resections, but several tumor
foci were not completely resectable. The patient was suffering from mainly two
tumor foci. First, the chordoma in the right pararectal space
(Fig. 1A) caused severe local
pain and a disturbing sensation of frequent urgent bowel movements. Second,
the chordoma near the left ischiadic nerve aroused persisting sciatic
neuralgia. In consensus with our surgeons, we suggested radiofrequency
treatment under MR guidance. The patient agreed to the attempt of
radiofrequency ablation.
View larger version (112K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A —66-year-old woman with recurrent sacrococcygeal chordomas.
A-C are T2-weighted transverse images. Follow-up images
D-F are T1-weighted fat-saturated contrast-enhanced with
high-field-strength MRI. For radiofrequency treatment, patient was placed in
prone position on left side. (These images have been rotated to resemble
regular supine scanning position.) Before procedure, location of paraanal
sacrococcygeal chordoma was determined (arrow, A). Using
multiplanar reconstruction in MRI, radiofrequency applicator (arrow,
B) could be exactly placed in center of tumor (only transverse section
is shown). After treatment, hyperintense signal of tumor is replaced by
hypointense coagulated material (arrow, C).
|
|
Before ablation, localization of the tumor foci was obtained with
T1-weighted spin-echo (TR/TE, 400/20) and T2-weighted turbo spin-echo
(4,080/109) imaging on a 2-T open scanner (Magnetom Concerto, Siemens Medical
Solutions), with and without application of 0.1 mmol/kg of body weight of
gadopentetate dimeglumine (Magnevist, Schering). Radiofrequency ablation
itself was performed using an MR-compatible radiofrequency system composed of
a 480-kHz radiofrequency generator (CC, Valleylab) and an internally cooled
MR-compatible applicator (CTM 1530, Valleylab). The length of the needle was
15 cm, including an active tip of 3 cm. Local anesthesia was achieved with
lidocaine 1%. The MR-compatible applicator was placed using MR fluoroscopy
(fast imaging with steady-state free precession [FISP], 50/18). For both
ablations, exact needle position was confirmed before ablation after
multiplanar reconstruction (Fig.
1B). Afterward, the treated regions were assessed using the same
preinterventional MR settings (Fig.
1C). The total duration of the procedure, including applicator
placement, was 90 minutes. Under CT guidance, the total time of the procedure
would have been similar because of the required time-intense computed
reconstructions of the intervention area.
View larger version (129K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B —66-year-old woman with recurrent sacrococcygeal chordomas.
A-C are T2-weighted transverse images. Follow-up images
D-F are T1-weighted fat-saturated contrast-enhanced with
high-field-strength MRI. For radiofrequency treatment, patient was placed in
prone position on left side. (These images have been rotated to resemble
regular supine scanning position.) Before procedure, location of paraanal
sacrococcygeal chordoma was determined (arrow, A). Using
multiplanar reconstruction in MRI, radiofrequency applicator (arrow,
B) could be exactly placed in center of tumor (only transverse section
is shown). After treatment, hyperintense signal of tumor is replaced by
hypointense coagulated material (arrow, C).
|
|
View larger version (112K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1C —66-year-old woman with recurrent sacrococcygeal chordomas.
A-C are T2-weighted transverse images. Follow-up images
D-F are T1-weighted fat-saturated contrast-enhanced with
high-field-strength MRI. For radiofrequency treatment, patient was placed in
prone position on left side. (These images have been rotated to resemble
regular supine scanning position.) Before procedure, location of paraanal
sacrococcygeal chordoma was determined (arrow, A). Using
multiplanar reconstruction in MRI, radiofrequency applicator (arrow,
B) could be exactly placed in center of tumor (only transverse section
is shown). After treatment, hyperintense signal of tumor is replaced by
hypointense coagulated material (arrow, C).
|
|
Results
Results of previous studies suggest that radiofrequency-treated lesions
have a characteristic appearance on contrast-enhanced T1- and T2-weighted MR
images [7]. On T2-weighted
cross sections, an inner hypointense zone corresponds to necrotic tissue, and
a hyperintense outer zone corresponds to edema or hemorrhage surrounding the
ablation area. Therefore, the first radiofrequency ablation was stopped after
2 minutes to acquire T2-weighted images for the assessment of signal changes
in the treated area. In accordance with these observations, viable chordoma
presented with hyperintensity on T2-weighted images, and coagulated tissue
became hypointense (Figs. 1B
and 1C). For radiofrequency
ablation of the paraischiadic chordoma, we delivered 39.9 kJ heat energy to
this area (130 W for 2 minutes [electric current1 = 1.3 A] and 135
W for 3 minutes [electric current2 = 1.35 A]), with intermittent
MRI (T2-weighted) after 2 minutes and 3 minutes, respectively. Subsequently,
49.2 kJ was applied to the paraanal tumor (164 W for 2 minutes [electric
current1 = 1.6 A] and 164 W for 3 minutes [electric
current2 = 1.64 A]), and intermittent MRI control was also obtained
after the first 2 minutes. Ablation areas showed clear loss of signal
intensity at T2-weighted imaging, corresponding to structural changes within
the tumors. Immediately after the radiofrequency session, the patient
experienced a complete regression of her symptoms. Two hours after
postinterventional monitoring, the patient refused the previously scheduled
admission because her complaints had become nonexistent, and she was
discharged.
View larger version (106K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1D —66-year-old woman with recurrent sacrococcygeal chordomas.
A-C are T2-weighted transverse images. Follow-up images
D-F are T1-weighted fat-saturated contrast-enhanced with
high-field-strength MRI. In longitudinal study 4 weeks after thermal ablation
therapy, pararectal ablation area (arrow) shows thermal-induced
necrosis and peripheral inflammation. Surrounding muscular structures show
slight alterations, but no damages such as abscesses can be observed.
|
|
View larger version (91K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1E —66-year-old woman with recurrent sacrococcygeal chordomas.
A-C are T2-weighted transverse images. Follow-up images
D-F are T1-weighted fat-saturated contrast-enhanced with
high-field-strength MRI. In second control 4 months after ablation, region
shows scarring and residual edema (arrow). Necrotic areas are no
longer detectable in pararectal region. Alterations of surrounding structures
are regressive in obturator muscle.
|
|
View larger version (96K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1F —66-year-old woman with recurrent sacrococcygeal chordomas.
A-C are T2-weighted transverse images. Follow-up images
D-F are T1-weighted fat-saturated contrast-enhanced with
high-field-strength MRI. Regeneration (arrow) continues after 10
months. No signs of tumor relapse and complete regression of clinical symptoms
are indicators of successful ablation.
|
|
In follow-up examinations, the pararectal ablation area presented with
increasing tissue scarring (Figs.
1D,
1E, and
1F). The lesion in the vicinity
of the ischiadic nerve showed the preceding signs of tissue reorganization
without edema, and the size of the coagulative area was shrinking as well. In
summary, both radiofrequency-ablated sacrococcygeal chordomas showed signs of
progressive scarring and size reduction.
Discussion
Sacrococcygeal chordoma is the most common primary malignant tumor of the
spine in adults excluding lymphoproliferative neoplasms
[8]. The therapy of choice is
surgical resection, but prognosis remains poor because sacrococcygeal
chordomas have an almost 100% recurrence rate, resulting in a 10-year average
survival.
In an interdisciplinary approach to the treatment of sacrococcygeal
chordomas, percutaneous radiofrequency ablation is an alternative to surgery,
which could be considered a new first choice in the treatment of recurring
mesenchymal tumors.
Our patient reported immediate relief of symptoms after the thermal
ablation. Radiofrequency treatment of the paraanal lesion has saved her from
proctectomy and diverting colostomy. But radiofrequency ablation in sensitive
locations like the paraproctium or in the vicinity of important vessels and
nerves can be accompanied by severe complications. In our case, perforation of
the rectum, lesion of the anal sphincter with anal incontinence, and definite
destruction of the ischiadic nerve were obvious risks. In the vicinity of
nerves, nerve conduction studies could help monitor the integrity of the
structure. But our patient was not sedated and thus was able to provide verbal
feedback.
Although the field of radiofrequency is constantly growing with regard to
the types of tumors that are potentially treatable with this method,
algorithms to monitor the course of the ablation do not exist for treatment of
tumors other than hepatic malignancies. MRI has some advantages over CT and
sonography for guidance of radiofrequency ablation of chordomas. First, the
nucleus pulposus-like tissue features of the mesenchymal sacrococcygeal
chordoma make MRI the technique of choice. T2-weighted imaging clearly
delineates chordoma from the surrounding tissue, allowing an excellent
assessment of the tumor. Second, MRI allows multiplanar reconstruction for
optimal spatial and temporal needle positioning into the tumor. Third,
interventional MR monitoring allowed us to assess immediately the extent of
thermally induced tissue alterations. The loss of protons in coagulated tissue
corresponds to a readily detectable and sensitive signal change in T2-weighted
imaging. Because no protocols are available for the radiofrequency ablation of
chordomas and many other tumors (in terms of power, current, time, and
resistance), the quick loss of signal intensity in T2-weighted imaging makes
controlled empirical procedures possible and safer. In our case, the
intraprocedural monitoring precisely directed us to interrupt the intervention
after a total ablation time of 5 minutes. In contrast, specific findings on CT
after radiofrequency of neoplasms are air bubbles and lack of signal
enhancement after the administration of contrast material. Unfortunately, the
amount of contrast material is limited because of its nephrotoxicity, and
after repeated injections of contrast material an equilibrium is established;
that is, further discrimination of the focus is not possible because of
residues of dye in the organ and in the lesion. Moreover, the guided probing
of the chordomas using CT fluoroscopy delivers high doses of radiation. At the
same time, the MR technique disqualifies standard metal radiofrequency
ablation probes for the intervention. Therefore, commercially available
MR-compatible probes were used and worked reliably.
In conclusion, despite lacking long-term data, this study shows that
MR-guided radiofrequency ablation is effective and safe and should be
considered for the treatment of malignancies located in the pelvis. Major
progress in the field of radiofrequency ablation depends on the development of
a noninvasive online monitoring of the induced thermal effects in patients.
Especially in areas with delicate anatomic structures, MRI of small regions of
interest may render radiofrequency ablation safer.
References
- Dupuy DE, Zagoria RJ, Akerley W, Mayo-Smith WW, Kavanagh PV, Safran
H. Percutaneous radiofrequency ablation of malignancies in the lung.
AJR 2000; 174:57
-59[Free Full Text]
- Callstrom MR, Charboneau JW, Goetz MP, et al. Painful metastases
involving bone: feasibility of percutaneous CT- and US-guided radio-frequency
ablation. Radiology 2002;224
: 87-97[Abstract/Free Full Text]
- Neeman Z, Patti JW, Wood BJ. Percutaneous radiofrequency ablation
of chordoma. AJR 2002;179
: 1330-1332[Free Full Text]
- Anis N, Chawki N, Antoine K. Use of radio-frequency ablation for
the palliative treatment of sacral chordoma. AJNR2004; 25:1589
-1591[Abstract/Free Full Text]
- McDannold NJ, King RL, Jolesz FA, Hynynen KH. Usefulness of MR
imaging-derived thermometry and dosimetry in determining the threshold for
tissue damage induced by thermal surgery in rabbits.
Radiology 2000;216
: 517-523[Abstract/Free Full Text]
- Lewin JS, Nour SG, Connell CF, et al. Phase II clinical trial of
interactive MR imaging-guided interstitial radiofrequency thermal ablation of
primary kidney tumors: initial experience. Radiology2004; 232:835
-845[Abstract/Free Full Text]
- Aschoff AJ, Rafie N, Jesberger JA, Duerk JL, Lewin JS. Thermal
lesion conspicuity following interstitial radiofrequency thermal tumor
ablation in humans: a comparison of STIR, turbo spin-echo T2-weighted, and
contrast-enhanced T1-weighted MR images at 0.2 T. J Magn Reson
Imaging 2000; 12:584
-589[CrossRef][Medline]
- Resnick DR, Greenway GD. Tumors and tumor-like lesions of bone:
imaging and pathology of specific lesions. In: Resnick DR, ed. Bone
and joint imaging. Philadelphia, PA: Saunders, 1996:1046
-1048
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
