Medical Dosimetry

Medical Dosimetry 39 (2014) 320–324

http://d0958-39

RepDirectiv

E-m

journal homepage: www.meddos.org

Treatment approach, delivery, and follow-up evaluation for cardiacrhythm disease management patients receiving radiation therapy:Retrospective physician surveys including chart reviews atnumerous centers

Michael S. Gossman, M.S., D.A.B.R., F.A.A.P.M.,* Jeffrey D. Wilkinson, Ph.D.,† andAvishek Mallick, Ph.D.‡

*Regulation Directive Medical Physics, Russell, KY; †Medtronic, Inc., Mounds View, MN; and ‡Department of Mathematics, Marshall University, Huntington,WV

A R T I C L E I N F O

Article history:Received 29 October 2013Received in revised form6 May 2014Accepted 6 May 2014

Keywords:CardiacDefibrillatorOncologyPacemakerRadiation

x.doi.org/10.1016/j.meddos.2014.05.00547/Copyright � 2014 American Association o

rint requests to: Michael S. Gossman, M.S., De Medical Physics, 104 Hildeen Court, Russellail: MGossman@TSRCC.com

A B S T R A C T

In a 2-part study, we first examined the results of 71 surveyed physicians who provided responses onhow they address the management of patients who maintained either a pacemaker or a defibrillatorduring radiation treatment. Second, a case review study is presented involving 112 medical recordsreviewed at 18 institutions to determine whether there was a change in the radiation prescription for thetreatment of the target cancer, the method of radiation delivery, or the method of radiation imageacquisition. Statistics are provided to illustrate the level of administrative policy; the level ofcommunication between radiation oncologists and heart specialists; American Joint Committee onCancer (AJCC) staging and classification; National Comprehensive Cancer Network (NCCN) guidelines;tumor site; patient's sex; patient's age; device type; manufacturer; live monitoring; and the reporteddecisions for planning, delivery, and imaging. This survey revealed that 37% of patient treatments wereconsidered for some sort of change in this regard, whereas 59% of patients were treated without regardto these alternatives when available. Only 3% of all patients were identified with an observable change inthe functionality of the device or patient status in comparison with 96% of patients with normal behaviorand operating devices. Documented changes in the patient's medical record included 1 device exhibitingfailure at 0.3-Gy dose, 1 device exhibiting increased sensor rate during dose delivery, 1 patient having anirregular heartbeat leading to device reprogramming, and 1 patient complained of twinging in the chestwall that resulted in a respiratory arrest. Although policies and procedures should directly involve thequalified medical physicist for technical supervision, their sufficient involvement was typically notrequested by most respondents. No treatment options were denied to any patient based on AJCC staging,classification, or NCCN practice standards.

& 2014 American Association of Medical Dosimetrists.

Introduction

Since the publication of the American Association of Physicistsin Medicine (AAPM) Task Group no. 34 (TG-34) in 1994, medicalphysicists and radiation oncologists have struggled with how tomanage radiation therapy for patients when heart rhythm devicesremain implanted during treatment.1 Such devices include

f Medical Dosimetrists

.A.B.R., F.A.A.P.M., Regulation, KY 41169.

implantable cardiac pacemakers (ICPs) and implantablecardioverter-defibrillators (ICDs). The evidence of this strugglewas exhibited in a survey published by Solan et al. in 2004. Theyhad statistically analyzed concerns and strategies for handlingpatients possessing either of these 2 devices during radiotherapyand also detailing the inconsistency in safety management forsuch patients among 71 responding ROs across the nation.2 TG-34guidelines for cardiac rhythm disease management (CRDM) inradiation oncology are now considered to be outdated. AAPM TG-203 has been charged with the effort to readdress all relatedrecommendation.1 TG-34 guidelines do not encompass modern

Fig. 1. Physician's awareness of society-based or manufacturer-based recom-mended radiation limits for CRDM devices.

M.S. Gossman et al. / Medical Dosimetry 39 (2014) 320–324 321

technology and perhaps more importantly do not deal with ICDs atall. Still, this document widely remains the current practicestandard within the community of ROs and medical physicists.

The intent of this research was to combine the outcomes of2 independent studies about how patients possessing cancer withimplanted CRDM devices are managed when dealing with theirneed for radiation therapy nearly a decade after the research bySolan et al.2 First, in a survey study described here, 200 randomlyselected physicians with either radiation oncology or vascular carespecialties were asked to respond to various questions provided.The study design sought to qualitatively and quantitatively eval-uate the number of patients seen by them annually, to assess theirfamiliarity with dose limits and adequate management, and toexplore differences in how patients were actually managed.Secondly, in a case review study described additionally, themedical records of 112 radiation therapy patients treated at 1 of18 institutions were retrospectively examined to determinewhether radiological concerns about damage to such devices weresignificant enough for a RO to invoke a change in the radiationprescription for the treatment of the targeted cancer, the methodof radiation delivery, or the method of radiation image acquisition.

Methods and Materials

Survey Study

To carry out the survey study, physician information was obtained throughaccess to the American Medical Association database. For this purpose, 200physicians from each of 4 different specialties: general cardiology, cardiac electro-physiology, cardiothoracic surgery, and radiation oncology were randomly selectedfrom around the United States. The ROs were considered as one group, whereasgeneral cardiologists, cardiac electrophysiologists, and cardiothoracic surgeons(CCC) were considered as a second group to quantify combined statistics. Themeans for analyzing the survey results were made possible by the use of a securedweb-based survey site contracted by Pro eBusiness, Inc. (Proctorville, OH). Requestswere sent by envelope mail and electronic mail identically. Each request wasinclusive of passwords for internet site access to complete the survey. Responsesreturned by envelope mail were transferred to the secured website manually by ourresearch team.

Statistics were provided to illustrate the level of administrative policy, proce-dure, and experience of the RO in managing such patients, as well as the level ofcommunication between ROs and the implanting heart specialist. Given the finitenumber of acceptable methods possible to perform such a study, we soughtimportance to design it specifically with our need to normalize results over theentire physician's practice rather than to 1 specific manufacturer, device, or model.As is common for surveys, only a limited number of those solicited chose torespond. In the results that are reported here, the number of usable responses isdenoted in parentheses following a value. Summary statistics are generallyreported as a median or the middle value and the range of the middle 50% of thevalues. This is known as the interquartile range and abbreviated “IQR.”

Case Review Study

For the case review study, 18 different institutions permitted through theirinstitutional review board a retrospective review of 112 medical records on patientshaving an intact CRDM device for an assessment of the treatment, imaging, andsafety management. Included in the survey were recommendations from theAmerican Joint Committee on Cancer for staging and classification and the NationalComprehensive Cancer Network guidelines for treatment delivery.3,4 In total, 12clinical considerations were also included in our investigations. These include thetumor site; patient's sex; patient's age; device type; manufacturer; model; theparticipation of cardiologists or manufacturer representatives; heart rhythm devicerevisions and interrogations; documented options for planning, delivery, imaging,and finally any observed clinical changes in the patient or the device as a result ofpotential radiation damage or interference from radiation.5,6

Results

Survey Study

In the survey study, of the 200 polled physicians, 71 respondedincluding 44 ROs and 27 cardiac physicians from the CCC group.The focus of the study was to ascertain the knowledge of radiation

safety guidance documents, such as that from AAPM TG-34 orfrom a specific device manufacturer. Therefore, the main questionsthat were asked were whether the physician was familiar withradiation limits for (1) the ICP and then for (2) the ICD. Of the 71physicians, 20 (28%) answered “No” to both questions, whereas 46(65%) answered “Yes” to both. Of the remaining 5 physicians, 3 said“Yes” to only ICP limits and 2 said, “Yes” to only ICD limits. This isbetter illustrated in Fig. 1.

Following up to the previous 2 questions, the RO group wereasked, “What is the likelihood, if any, that the dose to such a devicewill exceed the recommended limit?” All responded (100%) “Notlikely.” This response is noteworthy in light of the fact that in aprior question, 25% (11 of 44) of the RO respondents were notaware of radiation limits of the manufacturer or the nationallyaccepted society-based limits to either ICP or ICD devices.Respondents of the RO group were then asked, “Do you have yourqualified medical physicist review the dosimetry of such cases foraccuracy prior to treatment?” Even though not all ROs knew orcould reference the radiation limits on such devices when askedabout them specifically, each answered that they always refer to aqualified medical physicist to preview treatment for those casesinvolving ICP or ICD devices. This relieves some concern on thelack of safety knowledge, where the person responsible for allscientific and technical aspects of treatment is sought after forinvolvement and case awareness.

Differences in the responses between the CCC group and ROgroup may be related to the size of each physician's practice. Inreview, we advance to the responses of the RO group when asked,“In your practice, how many cancer patients were treated annu-ally?” and to the CCC group when asked, “How many CRDMdevices patients were seen annually in your practice?” For theRO group, the median was 800 patients (IQR: 300 to 1500,maximum ¼ 30,000, n ¼ 40), and for the CCC group it was 500patients (IQR: 200 to 800, maximum ¼ 3000, n ¼ 13). It is useful tonote that the 5 largest RO practices reported 2213 patients (n ¼ 1),5000 (3) and 30,000 patients (1) while the 5 largest CCC practicesreported 600 (1), 800 (1), 1000 (2) and 3000 (1) patients. Annually,the average number of patients in both groups for physicianssampled was substantial.

The RO clinicians were asked how many patients with CRDMdevices were seen annually in their practice that required a changein treatment approach owing to the presence of their device. Themedian number of patients was 4 (IQR: 6 to 8, maximum ¼ 35, n ¼44). Thus, the fraction of patients requiring a change in treatmentapproach annually was calculated as 0.8% (IQR: 0.4% to 1.5%,maximum ¼ 6.7%, n ¼ 40). Presumably, the total fraction of devicepatients seen each year is somewhat higher than this value asmany would not require a change in approach. This is consistent

M.S. Gossman et al. / Medical Dosimetry 39 (2014) 320–324322

with the findings of Solan et al.2 that “any busy radiation oncologydepartment should expect to face the prospect of treating a patientwith one of these devices at least several times annually.” Tounderstand the choices made for managing these patients the CCCclinicians were asked 2 questions regarding the number of timesper year that they surgically “revised” (either by reimplant orextraction) the implant before therapy. The median response forthe combined questions was 0.1 requests for revision surgery (IQR:0 to 1, n ¼ 15) with the median response at 0.033% (IQR: 0% to0.29%, n ¼ 11). The CCC group reported that 80% (IQR: 20% to 90%,n ¼ 6) of the requests received from a RO for surgical revision weresupported.

Similarly, the RO group was also asked a pair of questionsregarding surgical revision. For the patients that required somemodification to treatment owing to the presence of a CRDMdevice, 30% of patients (IQR: 10% to 50%, n ¼ 37) were subject torevision surgery. The RO group reported that 80% (IQR: 16% to100%, n ¼ 27) of requests for surgical revision were fulfilled. This isconsistent with the results tallied by the CCC survey. The groupwas also asked to report how many of these patients were deniedradiation therapy. Although the median was 0 (IQR: 0 to 0, n ¼ 40),there were a total of 6 respondents of 40 that reported denial ratesof 0.1 (n ¼ 4), 0.3 (1) and 0.4 (1) annually. These reports were fromsmall- to medium-sized clinics ranging in size from 250 to 900patients seen annually.

Case Review Study

For the case review study, 112 medical records were reviewedfrom 18 different institutions. The distribution of the devicemanufacturers reported included Medtronic at 55%, St. Jude at28%, Boston Scientific at 11%, Biotronik at 4%, and Vitatron andGuidant at 1%. Approximately 70% of the patients were betweenthe ages of 61 and 80 years and only 2 patients were younger than50 years. Almost two-thirds (n ¼ 76) of the patients were men. Theaverage age of men was 69.8 years with women at 71.9 years. Thedistribution of organ of treatment for all patients is illustrated inFig. 2.

It was observed that 63% (67, n ¼ 107) of the patients had anICP implanted, with the remainder having an ICD. As far as thecancer stage information of these patients with arrhythmia isconcerned, it was found that 30% (n ¼ 70) of patients had Stage I,19% had Stage II, 30% had Stage III, and 21% had Stage IV disease.Three-dimensional conformal radiation therapy (63%, n ¼ 80) wasthe most common type of radiation treatment delivered, followedby intensity-modulated radiation therapy (30%). It was found fromanalysis of the medical records that a cardiologist was notconsulted to participate in monitoring the patients at some point

Fig. 2. Tumor site of treatment for patients with CRDM devices intact (rounded).

during the radiation therapy in more than half (51%) of all cases.Among those involving a member of the CCC group, at times dailydevice interrogation was conducted both pretreatment and post-treatment, sometimes only before the first treatment delivery,and some during only 1 of the first 3 treatments. In few cases, thedevice was checked semiweekly or weekly. For most, vitalswere taken both pretreatment and posttreatment delivery for3 fractions with electrocardiogram monitoring conducted on thefirst 2 days.

Data in the medical records revealed that in many cases, acardiologist was not involved either to avoid hospital professionalcommunication issues or owing to the lack of knowledge of how acardiologist could help. In 7% (n ¼ 8) of cases, the devicemanufacturer was present for direct assistance although for onlythe first 1 to 3 fractions to check programming and operationalsettings. In a few cases (n ¼ 4), a magnet was used to put thedevice in standby during treatment delivery. However, only in 4%of all cases, the device was revised or reprogramed by a cardiol-ogist before radiation treatment on request by the supervisingphysician. The revisions recorded were noted for only a few cases(n ¼ 4) as follows: (1) a patient complained of an irregularheartbeat, was sent to the Emergency Department, where acardiologist interrogated the patient and then reset the device,(2) the RO requested revision in the placement of the device inrelation to a breast field, (3) the RO requested revision because ofthe placement of the device in relation to a lung field, and(4) device revision based on reprogramming issues.

Approximately 94% (n ¼ 98) of the patients were treatedaccording to the 19-year-old TG-34 report guidelines (AAPM).In the remaining cases, the manufacturer's limits were used. Forsome facilities, the clinic policy was to adhere to the more strictlimit between the AAPM and the manufacturer. Approximately in85% of the cases, the CRDM device received less than 2 Gy,whereas the remaining 15% received more as illustrated in Fig. 3.

Approximately 14% of medical records revealed considerationfor neutron interactions induced when higher energy x-rays areused (48.04 MV). For these cases, multiple x-ray energies wereclinically available, and the lower energy was used. For all casesobserved, the lesser energy was 6 MV. Medical records indicatedno consideration for neutron effects at all in 75% of the recordsreviewed.

In 84% of cases, no precautions were taken on reducing thedose rate of the particle accelerator when reduction was possible.Dose rate was simply not considered an issue from medical recordreview and queries to the center. Only in 4% of the cases was thedose rate considered and yet increased anyway. The remaining 12%of cases involved treatments using machines with only a singlerate of delivery.

Fig. 3. Maximum dose given to the CRDM device during treatment.

Fig. 4. Considerations given by the radiation oncologist to reduce dose, energy, ordose rate. Fig. 5. Willingness of the radiation oncologist to provide treatment.

M.S. Gossman et al. / Medical Dosimetry 39 (2014) 320–324 323

In 71% of the cases, using daily imaging methods with lower x-ray intensity was not at all considered. In 25% of cases, alternativeimaging options were reviewed with changes such as adjustingMV cone-beam computer tomography (CBCT) daily to weekly,avoiding kV CBCT, changing kV port daily to MV port weekly,changing CBCT to kV-kV port matching, or opting to acquire MVports during actual treatments rather than adding independentports. In 4% cases, patients were treated with no required imagingneed. An example of these cases may be electron treatments to theskin (3%) or a simple beam setup to an extremity (1%). Overall, only37% of patient treatments were considered for some sort ofchange, whether it was an alternative energy, dose rate, or imagingoption, whereas 59% of patients were treated without regard tothese alternatives when available (Fig. 4).

In only 3% of patients, observable changes in device function-ality were concluded to be the direct result of radiation treatment.Medical record documented changes included 1 device exhibitingfailure at a 0.3 Gy dose, and 1 device exhibiting increased sensorrate during dose delivery. Clinical changes for patients included1 patient having an irregular heartbeat leading to device reprog-ramming from a CCC member in the Emergency Department.Another patient complained of twinging in the chest wall andnear the device position for last few remaining treatments, whichresulted in a respiratory arrest. In 96% of the cases, patients wereidentified with normal behavior and normal operating devices.

Discussions

Considering the relatively small proportion of a physician's totalpatients that require radiation therapy with a CRDM deviceimplanted, it is difficult to gather reliable data on these eventsby means of a retrospective survey. The standard methodology inthe study of rare events is a process known as “oversampling” inwhich a very large sample is taken in the hopes of producing alarge enough sample of the rare event to discover its traits. Giventhe large patient base of the RO group, there was distinctly morelikelihood in identifying the traits of this rare event within thatdata set. Given this observation, the RO group was asked if theywould be less willing to prescribe radiation therapy to the patientswith an implanted CRDM device. Responses from 13 (29.5%, n ¼44) RO physicians indicated that they are less willing to prescriberadiation therapy to such patients and their concerns were focusedto removal of risk foremost, then the consequences of theirpatients suffering through multiple procedures, but not infectionrisk at all. This is detailed in Fig. 5.

Probing further to the 31 ROs who indicated they were not lesswilling to provide treatment, each were then asked: “What else

might change your willingness to prescribe radiation therapy?”Open responses included 26% who reported patient reluctance,10% deferred until after a consult could be obtained with thecardiac physician specialist, and 10% were concerned about devicesbeing directly in the field of radiation. When the RO group wasasked about an increase in morbidity or mortality when the devicewas moved or removed, 33 of the 44 replied “No.” One physiciancited experience with infection resulting in device removal.

Similar results came from the 24 members of the CCC group.One reported marginally increased risk of infection and anotherreferenced acceptable clinical risks. When the RO group wasasked: “Does your standard practice ever deviate depending onthe immediate care need of your patient?” A total of 9 respondentschose “Yes,” with additional open responses for 5 who stated thatthey would monitor the CRDM, monitor the patient, and requestin vivo dosimetry from their qualified medical physicist. Only1 respondent indicated that they would modify the beam entranceangle to limit the dose to the heart rhythm device. This method-ology was suggested by Gossman7 noting that such devices shouldbe a few centimeters out from the radiation field edge to avoid upto 50% penumbral dose. Treatment planning system inaccuracieswere also explicitly detailed for the calculation of dose fromexternal beam radiation.8,9 Possible consequences of radiationdamage to the patient or device from radiation were providedadditionally from the experience of all respondents. These include3 reports of ICD shock; 2 reports of lightheadedness and decreasedblood pressure; and 3 reports indicating ICP resets, general resets,or erased memory. Another indicated an ICP patient experiencedpermanent operational damage, which required replacement.

Results differentiated according to the size of practice werenearly identical within the CCC group. However, results concerningRO group responses were remarkable for their size of practice. Theaverage number of times a device was moved or extracted per yearwas reported as 0.0012 (0.12%). Of the 39 physicians responding,20 had an individual ratio less than 0.0012 (0.12%), whereas 19 hada greater than 0.0012 ratio. One interesting observation was thatindividual physicians with a small to average patient load tend tohave a higher ratio of patients who had the device moved orremoved compared with those physicians having a larger practice.It can be inferred then that smaller to average size RO practices tendto request a revision to the CCC physician more often than the largercenters. As an example, one might suppose here that RO's from astand-alone cancer center are more likely than a university-basedcancer center to have significant communication with a CCC doctorfor the purposes of considering relocation strategies.

A challenging aspect of the second data set was obtaining ananswer to every question desired. For example, some institutionsdid well to document the functional integrity of the device,

M.S. Gossman et al. / Medical Dosimetry 39 (2014) 320–324324

whereas others did not. At times, even though a cardiologist or arepresentative for the manufacturer was present during treatment,their results from electrocardiogram, interrogation, programming,settings for voltage, etc. were not provided in the medical record.

There were unidentified device models for some patients aswell. For these patients, manufacturer identified 94% of devices. Itwas found that manufacturers each have a Patient RegistrationDivision (Medtronic), Patient Records Division (St. Jude), or MedicalRecords Division (Boston Scientific) who confirm the device type(ICP/ICD), model number, and serial number if the patient's name isprovided along with their date of birth. The result of knowingwhich device specifically the patient was implanted with did showan indication of consideration to deviate from the recommenda-tions of the 19-year-old TG-34 publication. Approximately 94% ofpatients were treated with emphasis on the TG-34 publishedguidance, and in the remaining cases, the manufacturer's limitwas used. However, a device limiting assignment on dose to 2 Gywas only followed for 91% of all patients. The remaining 9% plannedfor a device tolerance of 2 to 10 Gy. Approximately 63% of thedevices were pacemakers. Given that fact, it is possible that theremaining 37% of defibrillator cases were not applicable to TG-34,as it did not address that device type. Hence, ROs and medicalphysicists were left with no alternative but to adopt TG-34 fordefibrillators too or to make a call and assign the manufacturer'slimits. The lack of certainty and therefore a mark of patternedinconsistency is that one would then expect 63% of cases to havebeen assigned a TG-34 device limit of 2 Gy as they were pace-makers, vs the 91% found. Approximately 85% of cases resulted in adose to the device of less than 2 Gy, whereas 15% of devicesreceived more than the AAPM recommended limit.5

Shortcomings in the research were made apparent in bothsurvey methods. In the survey study, 800 requests were sent out,both via paper mail and via email with secure passwords forwebsite participation. A total of 200 invitations were sent out tophysicians within each of the 4 specialties. Only 71 physiciansresponded. Still, the tally was appropriately varied in both the CCCgroup and the RO group across the U.S. regionally. The size of thepractice was also reviewed. We noted that physicians with small-to medium-sized practices mostly provided the data. A total of 112medical records were reviewed at 18 institutions. Some centersprovided multiple medical records, whereas others had only 1 or 2.To reduce the time and cost burden for long-distance journeys,much of this research was conducted by centers in the midwest,southeast, and mid-Atlantic regions of the United States. A broadreach for this study was never sought after or achieved. However,the data set does provide a relatively even distribution ofresponses from university-affiliated centers and smaller clinics.

Conclusions

Most ROs surveyed do not adhere to recommended standardsof practice by the AAPM, primarily as a result of unfamiliarity.Although policies and procedures should directly involve thequalified medical physicist for technical supervision and radiationtreatment management, they were found to be insufficientlyinvolved a priori.2,5

No treatment options were denied to any patient based onAmerican Joint Committee on Cancer staging and classification, andno deviations from the recommended National ComprehensiveCancer Network practice standards were identified for any patientas a consequence of the patient having an intact heart rhythmdevice.3,4 Machine options often permitted a change to use

alternative energies, dose rates, or imaging modalities. However,this survey revealed that only 37% of patient treatments wereconsidered for some sort of change in this regard, whereas 59% ofpatients were treated without regard to these alternatives whenavailable. For centers considering any of these, 14% reduced thetreatment energy, 4% reduced the dose rate, and 25% changedimaging verification methods.

Only 3% of patients were identified with an observable changein the functionality of the device or patient, in comparison with96% of patients with normal behavior and normal operatingdevices. Disease staging, classification, device type, treatmentmethodology, and treatment dose were unaltered as a result ofthe patient maintaining an implantable heart rhythm device. Withstrategic consideration of device susceptibility to damage at 37%,the survey projects a need for education and guidance on theappropriate means for radiation treatment and image acquisitionwhen caring for patients with arrhythmia.

To date, appropriate communication pathways between theheart rhythm specialist and RO are substantially unexploited, eventhough patients requiring radiation therapy with a deviceimplanted were significantly more clinically challenging. Expand-ing these opportunities with bilateral education and scientificawareness promotes a better understanding of device limitationfor each, while offering both a more accurate prospectus whenconsulting patients who are seeking information about suchconcerns. This research indicates that professional educationshould be progressively promulgated for each of these physicianspecialties when managing patients who require radiation treat-ment for cancer care while continuing treatment for heart rhythmabnormalities as continually recommended by the AAPM and theAmerican Association for Therapeutic Radiology and Oncology.5,6

Acknowledgments

The opportunity to undertake this large-scale survey wassupported by the issuance of Medtronic, Inc. grant nos. 1921 and1390 (to M.S.G.). The authors thank all of the radiation oncologistsand medical physicists from clinics nationwide who assisted withinstitutional review board applications and facility permissions togain privileges for audits and the compilation of these census data.

References

1. Marbach, J.R.; Sontag, M.R.; Van Dyk, J.; et al. AAPM Report no. 45 (TG-34)—Management of radiation oncology patients with implanted cardiac pacemakers.Med. Phys. 21:85–90; 1994.

2. Solan, A.N.; Solan, M.J.; Bednarz, G.; et al. Treatment of patients with cardiacpacemakers and implantable cardioverer-defibrillators during radiotherapy. Int.J. Radiat. Oncol. Biol. Phys. 59:897–904; 2004.

3. American Joint Committee on Cancer (AJCC) cancer staging handbook, 6th ed.,Springer-Verlag, New York, 2002.

4. National Comprehensive Cancer Network (NCCN) Guidelines.5. Marbach, J.R. Recommended precautions in the management of radiation

oncology patients with implanted pacemakers. ASTRO Newsletter VII; 1989:7–8.6. White, G.A. Letter to the U.S. FDA. Center for Devices and Radiological Health; 21

April 2008.7. Gossman, M.S. Chapter 9: Clinical concerns for and strategies with pacemakers

in radiation oncology. In: Vonend, O., Eckert, S., editors. Aspects of pacemakers—Functions and interactions in cardiac and non-cardiac indications, pacemakers—Book 2. Rijeka, Croatia: InTech. 145–62; 2011.

8. Gossman, M.S.; Graves-Calhoun, A.R.; Wilkinson, J.D. Establishing radiationtherapy treatment planning effects involving implantable pacemakers andimplantable cardioverter-defibrillators. J. Appl. Clin. Med. Phys. 11:1–13; 2010.

9. Gossman, M.S. Radiation oncology dose gradients induced by heart rhythmdevices. Diagn. Intervent. Cardiol. Mag. Adv. Radiat. Oncol., 1–5; 2011.