Abstract-The medical community is advocating for progressive improvement in the design of implantable cardioverter-defibrillators and implantable pacemakers to accommodate elevations in dose limitation criteria. With advancement already made for magnetic resonance imaging compatibility in some, a greater need is present to inform the radiation oncologist and medical physicist regarding treatment planning beam profile changes when such devices are in the field of a therapeutic radiation beam. Treatment plan modeling was conducted to simulate effects induced by Medtronic, Inc.-manufactured devices on therapeutic radiation beams. As a continuation of grant-supported research, we show that radial and transverse open beam profiles of a medical accelerator were altered when compared with profiles resulting when implantable pacemakers and cardioverter-defibrillators are placed directly in the beam. Results are markedly different between the 2 devices in the axial plane and the sagittal planes. Vast differences are also presented for the therapeutic beams at 6-MV and 18-MV x-ray energies. Maximum changes in percentage depth dose are observed for the implantable cardioverter-defibrillator as 9.3% at 6 MV and 10.1% at 18 MV, with worst distance to agreement of isodose lines at 2.3 cm and 1.3 cm, respectively. For the implantable pacemaker, the maximum changes in percentage depth dose were observed as 10.7% at 6 MV and 6.9% at 18 MV, with worst distance to agreement of isodose lines at 2.5 cm and 1.9 cm, respectively. No differences were discernible for the defibrillation leads and the pacing lead. 2011 American Association of Medical Dosimetrists.
The first pacemaker was designed by John A. Hopps in 1949. From efforts to follow by investigators, such as the 1961 Nobel Peace Prize recipient, Bernard Lown, and the team of Michel Mirowski and Morton Mower, the first ever implantable cardioverter-defibrillators (ICD) and implantable pacemakers (IPs) became possible 20 years after Hopps' artificial invention. These devices are now well-known, highly recommended, and routinely implanted by electrophysiologists and cardiovascular physicians internationally. As with all electronic devices, there are limitations to their usefulness. Clinically, the device may at times present a limitation for other medical specialists. One such area is in the field of radiation oncology.
In vitro testing of devices in radiation beams have led to findings of stability and functionality changes. With the consideration to those inherent operational deviations, radiation oncologists are apprehensive about possible effects that may be detrimental to the patient if a device remained implanted and used during or after radiation therapy. This is important even for the initial computerized tomography (CT) image acquisition, from which diagnostic staging and treatment modeling is based. Investigators hope to identify the electronic effects revealed by such cardiovascular devices through research already underway. The results of such investigational work may shed new light on recommended dose limitations, whereby modern devices with radiation-hardened electronics and magnetic resonance imaging (MRI) compatibility provide greater potential capability for sustaining functionality at doses exceeding the American Association of Physicists in Medicine (AAPM) recommendations of under 2 Gy. With the AAPM advocating to the United States Food and Drug Administration to assist in the promotion of education and training on these devices in radiation oncology, direct concerns regarding the affects of such devices on a therapeutic beam remain unpublished. As the functional changes in the electronic devices are being examined elsewhere, we present a focus study on the effects of radiation therapy beam profile shifts though computerized modeling.
The grant-supported research compiled here presents data for the radiation oncologist and medical physicist to reference, as a guide to treatment planning for cancer patients who require radiation when the device is still intact and intended for use during radiation delivery or afterward. Attenuation, side scatter, and backscatter have already been presented in the literature by this principal investigator under key research grant support. In that study, nearly identical physical effects were demonstrated for all implantable pacemakers. Indistinguishable phenomena were observed for all ICDs. It then follows that only one device from each group is necessary for study here. The required data presented is attentive to the magnitude and direction of shifts in therapeutic isodose depth dose curves caused by the introduction of an ICD and an IP. Axial and transverse beam profiles are analyzed at x-ray energies of 6 MV and 18 MV from a commercially available particle accelerator used to treat upwards of 60 patients per day for cancer therapy. This research constitutes results for more than two-thirds of all device families marketed by Medtronic, Inc.
A phantom was used to simulate a patient having a cardioverter-defibrillator or pacemaker implanted. The CNMC Company, Inc. (Nashville, TN) model WP-3040 water phantom was used for this purpose. The 12-gallon tank with longest dimension 40 cm defines the scanning volume for all CT data used. Acrylic plates were added to the bottom of the tank to provide a platform for devices to rest, while insuring adequate backscatter from the megavoltage x-rays to be modeled. Square plates 24.8 cm wide were stacked to a height of 8.0 cm inside the WP-3040 water phantom. The tank was then filled to raise the water level 5 .6 cm above the surface of the acrylic platform. This depth is also adequate to achieve build-up of dose at both 6 MV and 18 MV therapeutic x-ray energies. The ICD or IP was affixed to the center of the acrylic platform consecutively for each CT scan. Identical geometry was maintained in the experimental setup for each.
The General Electric (Fairfield, CT) Lightspeed RT scanner provided CT acquisition data. Scanning commenced after the programming of the helical mode stereotactic radiosurgery protocol: 120 kVp at 278 mA and an 80-second nominal scan time with 1.25 mm couch increments. The largest field of view at 5 0 cm was used. This process was repeated for both devices included in the study.
The Medtronic, Inc. devices include the biventricular ICD Concerto model C15 4DWK (VVE-DDDR) using defibrillation lead model 6947, and the IP Versa model VEDR01 using single pacemaker lead model 5076. The ICD Concerto and IP Versa generators are illustrated in Fig. 1. The Concerto ICD is the larger of the two. It has an orthogonal face area of 6.9 x 5.1 cm and thickness of 1.5 cm. The Versa IP is 3.1 times smaller in volume and 1.6 times smaller in area, with orthogonal face dimensions of 4.5 x 4.8 cm and thickness of 0.8 cm.
Scan image acquisition was commissioned for use of the extended Hounsfield units (HU) range. As discussed for published studies involving deep brain neurostimulator lead localization techniques, vascular access port imaging, and beam modeling, the extended HU ranges are important for the observance of high-density materials for submillimeter positioning accuracy specificity in neurosurgery, as well as for proper dose estimations in radiation oncology. The default range for the GE Lightspeed RT scanner is -1024 to +3071 HU. This differs substantially from its extended range of -31,743 to +31,743 HU. With the ICD found to contain iron, silver, and vanadium in the battery, and with aluminum in the high voltage capacitor, values of 3800 HU and 3000 HU result, respectfully. Although the IP device is smaller in all dimensions, the battery is composed of mainly iodine and iron, thus 8000 HU result for it. In addition, for the IP, a copper telemetry antenna is included in the design. The antenna alone yields 17,500 HU on average. These high-density areas are of focus for large isodose beam profile change locations, because doses will be computed with more attenuation and scatter for these high-density objects. Once each scan was reconstructed, the independent scan sets containing a total of 446 slices were transferred to a computer for dose calculation.
Dose delivery modeling was performed using the Varian Medical Systems, Inc. (Palo Alto, CA) Eclipse build version 8.6 external beam planning software. Artifacts were identified around each device, which occur as a consequence to improper sampling of attenuation data, from beam hardening of the 120-kVp CT beam through metal. With the beam incident only from the transverse direction, it then follows consistently with our findings that data streaking follows mainly lateral paths. Dose computation in treatment planning with highgradient false artifacts can result in miscalculation. Therefore, methods were suitably assigned to negate these observances.
Using existing knowledge of the dimensions of the ICD and IP, the software was used to carefully contour the surface area of each device on all CT slices. Then, a new structure was created using a Boolean operator, such that the new structure was identical in shape as the device scanned. A margin for the structure was programmed to increase its size by 3.5 cm. The new structure volume was then redefined using the Boolean tool, such that the resulting volume did not include the volume of the device being studied. The final new structure volume envelopes the ICD or IP device entirely with a margin of 3.5 cm and excludes the volume of the studied device within. An assignment of 0 HU for water density to this new structure, which encompassed the artifacts revealed in the CT dataset, ensured no analysis errors for dose computation.
. . .Continue to read rest of article (PDF).