Our institutional review board approved this retrospective study. From June 2008 to Mar 2011, 77 consecutive patients (Men: 45, Women: 32, mean age: 60 years old, range: 36-79) who had undergone percutaneous ultrasound-guided fiducial marker implantations for CyberKnife therapy (Accuray, Inc., Sunnyvale, CA, USA) of the intra-abdominal organs were enrolled in this study. All the patients elected to undergo CyberKnife radiation therapy after having a discussion with a radiologist, radiation oncologist, and oncologist. We implanted 270 fiducial markers into 77 patients (number of fiducials: two to six fiducial implantations/patient mean: 3.4 in the prostate, 3.5 in other sites including). Among the 270 implantations, 104 were inserted in the metastatic lymph nodes (32 patients), 96 were inserted in the liver (24 patients), 39 were inserted in the pancreas (12 patients), and 31 were inserted in the prostate (nine patients). The mean size of the target tumor had a maximum diameter of 3.3 cm.

Two-hundred twenty-two fiducial markers in 64 patients were inserted via the trans-abdominal route whereas 48 fiducial markers in 13 patients were inserted via the trans-rectal route. 31 fiducial markers in the prostate glands of nine patients, and 17 fiducial markers in the metastatic lymph nodes in the pelvic cavity of four patients were implanted under trans-rectal ultrasound guidance (Table 1).

All the fiducial marker implantations were performed by a gastrointestinal radiologist with 13 years of experience. By reviewing the previous diagnostic imaging studies for methods such as computed tomography, the best percutaneous needle approach was planned and fiducial insertion was performed under ultrasonographic guidance (IU 22, Philips, Einthoven, Netherlands) with a convex probe (2-5 MHz) at transabdominal ultrasonography and an endorectal probe (5-9 MHz) at transrectal ultrasonography. A free-hand style was used for the transabdominal ultrasonography, and needle guidance was used for the transrectal ultrasonography.

The skin entry site, which was kept sterile following disinfection with betadin, was prepared, and local anesthesia (lidocaine 1%, 10 mL) was administered. Under Ultrasound (US)-guidance the radiologist placed two to six gold pins (A fiducial marker kit, Civco, Maarn, the Netherlands) on and around the target, using 18-gauge introducer needles. Each fiducial marker measured 0.9 × 3.0 mm in diameter and length, respectively. When performing the fiducial implantation, it is advised to keep a minimum of 20 mm spacing, and a minimum 15° angle between the fiducials. Moreover, the fiducials should not be more than 50 to 60 mm away from the target.

This implantation procedure was usually performed within 20 minutes. Two or three days after the implantation, a CyberKnife therapy planning radiograph was performed, followed by a CyberKnife planning CT for the simulation of radiation treatment in all the patients within a span of two weeks. CT was performed using a 4-slice MDCT scanner (Sensation 4, Siemens Medical System, Forchhein, Germany). The imaging parameters included a 4.0 × 2.5 mm section detector collimation, 120 kV, 145 mAs, 15 mm per second table speed, pitch of 6, and a 1 mm reconstruction interval.

Upon review of the medical records CyberKnife therapy planning radiographs and the planning CT for the evaluation of procedure-related complications were performed. During and after the implantation, the major and minor procedure-related complications were documented. We only recorded the immediate complications until the time of the CyberKnife planning CT (within 2 weeks).

The major/minor complications were divided by the outcome and according to the SIR Classification System for Complications. Minor complications were divided into two types (A. No therapy and of no consequence, B. Nominal therapy and of no consequence; includes overnight admission for observation only). Major complications were divided into four types (C. Required therapy and minor hospitalization (< 48 hours), D. Required major therapy and an unplanned increase in the level of care, prolonged hospitalization (> 48 hours), E. Permanent adverse sequelae, F. Death).

Also, research about whether the fiducial implantation was successful or not was carried out, and we reviewed the CyberKnife therapy planning radiographs and planning CT images. We defined technical success as the implantation enabled adequate treatment planning and CT simulation. The cases in which more than two fiducial markers overlapped, or reimplantation was needed due to an unexpected occurrence such as fiducial migration, were considered as technical failure.

When conducting 270 fiducial implantations in 77 patients, there was one patient with major complication (1.3%) and 16 patients with minor complications (20.8%) (Table 2).

The patient with a major complication had previously undergone Whipple's operation due to CBD cancer. A newly-appeared, 3.7-cm, low-density mass in segment V of the liver on the follow up CT scan (this was presumed to be metastasis) was found. The patient was implanted with four fiducial markers under trans-abdominal ultrasound-guidance for the planning of CyberKnife therapy. Following the implantation, the patient complained of local pain in the insertion site. Analgesics and conservative treatment were performed; however, after 11 hours from the time of fiducial marker insertion, the patient's body temperature rose to 38.6℃ and dyspnea as well as right flank pain developed. The patient's pleural effusion gradually increased during the five days after the procedure. Following this, percutaneous-catheter insertion was performed for the pleural effusion. The patient's condition gradually improved after percutaneous-catheter insertion. The CyberKnife planning CT was taken two weeks after fiducial implantation and the CT scan showed a newly appeared subcapsular fluid collection around the liver and biliary dilatation of the peripheral intrahepatic duct, which was suspected to be a biloma. However, the patient did not complain about further symptoms. The patient suffered from a few complications, yet the four fiducial markers were well located in the target tumor. It could be seen on the follow-up CT that the tumor decreased in size from about 3.7 cm to 2.0 cm at one month after the CyberKnife treatment (Fig. 1).

Sixteen patients (20.8%) had minor complications that only required overnight admission for observation and no or minimal therapy. Eleven patients had abdominal pain, four of which did not require specific treatment during the follow-up period. Moreover, two patients had mild fever that also did not require any follow-up treatment, and one patient had mild transient bullae change at the skin puncture site and did not require specific treatment other than a simple dressing. Lastly, there were two patients with nonspecific urinary symptoms.

As guides for transrectal ultrasound, fiducial markers were inserted into fourteen patients. Among them, nine patients' fiducial markers were inserted to the prostate gland, and the other five into the metastatic lymph nodes in the pelvic cavity. When these implantations were conducted, one patient had transient hematuria that did not last for more than a day and it did not require any specific treatment. Another patient had transient voiding difficulty that required nelaton catheter voiding. None of the patients had major complications that required long-term hospitalization.

Of the 270 fiducial implantations, 261 (96.7%) were considered to be successful. However, the migration of fiducial markers was observed for six implantations (2.2%) in six patients. In the case of the planning CT scanning for the actual treatment, since the treatment is not feasible when fiducial markers migrate more than 6 cm from the target, the procedure needs to be re-done, which was considered a technical failure in this study. Of the six failed implantations, five were implanted for intraabdominal metastatic lymph nodes and one of them was implanted for hepatocellular carcinoma (Fig. 2).

We also observed a failure to discriminate fiducial markers for three implantations (1.1%). Each was implanted for an intra-abdominal metastatic lymph node (Fig. 3), cholangiocarcinoma and pancreatic body cancer, respectively. Among total 77 patients, six patients (7.8%) required additional procedures for reimplantation of fiducial markers, and one of them was reimplanted under a CT-guidance (Table 3).