S. Stintzing et al.
Here we compare treatment efﬁcacy and toxicityof the widely used percutaneous, computed tomog-raphy (CT)-guided, RFA and RRS for salvage treat-ment of colorectal liver metastases.
This analysis included patients that were treatedeither with percutaneous RFA or RRS therapy forcolorectal liver metastases. Hepatic lesions had to bedeemed surgical unresectable by the multidisci-plinary gastrointestinal tumor board of the Univer-sity Hospital Grosshadern, University of Munich.This board includes experienced hepatobiliary sur-geons, radiation oncologists, medical oncologists andinterventional radiologists. All metastases had to betreated in curative intent. Therefore, no extrahepaticdisease and no other lesions to the liver than thoseto be treated with local therapy had to be detectableprior to therapy. The decision if percutaneous RFAor RRS should be applied was based on the followingalgorithm: the standard procedure to treat unresect-able colorectal liver metastases was RFA. RRSwas preferred if one of the following reasons werepresent: vicinity of the metastasis to great vessels, asthose might cause a heat sink effect, location close tothe liver capsule, as RFA is more painful to thosepatients and patient wish. A maximum diameter of the lesion exceeding 30 mm has been a factor favor-able for the use of RRS during the last two years of the study .
Robotic radiosurgery (RRS)
Thirty consecutive patients with one to three livermetastases from colorectal cancer were treated incurative intent with single session RRS at theEuropean Cyberknife Center Munich as describedbefore [8,10] from May 2005 to May 2011. In short,a radiosurgical device (Cyberknife
, Accuray Inc.CA, USA) with real time tumor tracking was used.Prior to radiation, one or two gold ﬁducials (CPMedical Inc., Portland, OR, USA) were placed insideor next to the metastases via an 18G needle underlocal anesthesia and CT ﬂuoroscopy guidance in theinterventional radiology unit. Contrast enhanceddynamic CT (arterial and portalvenous phase) ormagnetic resonance imaging (MRI) (using liver-speciﬁc contrast agents) was used to exclude addi-tional intrahepatic tumors. A dedicated liver MRIexamination either performed with a Gadolinium-based extracellular [Magnevist
(GadopentetateDimeglumine) or Gadovist
(Gadobutrol), BayerHealthCare Pharmaceuticals Diagnostic Imaging,Germany] or a liver cell-speciﬁc [Primovist
(GadoxeticAcid), Bayer HealthCare Pharmaceuticals DiagnosticImaging, Germany] contrast agent, and character-ized by dynamic contrast-enhanced T1 weighted 3DFLASH sequences (Volumetric Interpolated Breath-hold Examination; VIBE; Siemens Healthcare, Erlan-gen, Germany) with a slice thickness of 3 mm wasfused to the planning CT in all cases by means of rigid image registration provided by the system soft-ware (Multiplan, Accuray Inc.). It was aimed to scanthe patient in the same breathing phase during CTand MRI to get a fusion result as good as possible.In difﬁcult cases the fusion was directed to the lesionitself and not to the outer margins of the liver. Addi-tionally, a margin of 6 mm was added to the tumordiameter in all three dimensions to accommodate fortumor cell spread around the tumor. The set-up inac-curacy for the RRS system used in this analyses is
1 mm . In a single session treatment, a totalof 24 – 26 Gy to the 70% isodose (dose maximum34 – 37 Gy) were administered to the target volume,using a 6 MV compact linear accelerator (LINAC)mounted on a six-axis robotic manipulator. Allpatients were pretreated with ondansetrone onthe day of treatment to prevent radiation-inducednausea.During treatment, respiratory motion tracking of the target volume was done as previously described. The whole procedure lasted about one hour,and patients were discharged from the instituteimmediately after treatment.
Radiofrequency ablation (RFA)
RFA treatment was performed by interventionalradiologists with at least ﬁve years of experienceunder single-shot antibiotics, pulse oximetry andconscious sedation using midazolam, dipyrone andpethidine or piritramide, respectively. All RFA pro-cedures were done on a 4-, 16-, or 128-row CT scan-ner (Siemens Healthcare, Erlangen, Germany) underCT ﬂuoroscopy guidance with angular beam modu-lation (tube voltage, 120 kV; tube current – time prod-uct per reconstructed image, 15 – 25 mAs). The lesionswere to be treated in a single session. If tumor sizewas too big, a second treatment was planned for thefollowing day. The subsequent RFA systems wereused: 1) AngioDynamics/RITA Medical Systems(Queensbury, NY, USA): 1500X RF Generator(max. 250W of power) with StarBurst
XL RFelectrode (umbrella diameter: 5 cm) and 2) BostonScientiﬁc (Natick, MA, USA): RF3000
Radio-frequency Ablation System (max. 200W of power)with LeVeen
RF electrodes (umbrella diameter: 2 – 5cm). After RFA treatment, patients were sent backto the ward for further pain management and clinicalmonitoring, and usually discharged from the hospitalafter CT follow-up 24 hours later.
A c t a O n c o l D o w n l o a d e d f r o m i n f o r m a h e a l t h c a r e . c o m b y U n i v e r s i t y o f S o u t h e r n C a l i f o r n i a o n 0 2 / 1 4 / 1 3 F o r p e r s o n a l u s e o n l y .