Int. J. Radiation Oncology Biol. Phys., Vol. 73, No. 5, pp. 1525–1531, 2009Copyright � 2009 Elsevier Inc.

Printed in the USA. All rights reserved0360-3016/09/$–see front matter

doi:10.1016/j.ijrobp.2008.06.1903

CLINICAL INVESTIGATION Implanted Defibrillator

IMPLANTED CARDIAC DEFIBRILLATOR CARE IN RADIATION ONCOLOGYPATIENT POPULATION

DAPHNA Y. GELBLUM, M.D.,* AND HOWARD AMOLS, PH.D.y

Departments of *Radiation Oncology and yMedical Physics, Memorial Sloan-Kettering Cancer Center, New York, NY

Purpose: To review the experience of a large cancer center with radiotherapy (RT) patients bearing implantablecardiac defibrillators (ICDs) to propose some preliminary care guidelines as we learn more about the devices andtheir interaction with the therapeutic radiation environment.Methods and Materials: We collected data on patients with implanted ICDs treated with RT during a 2.5-yearperiod at any of the five Memorial Sloan-Kettering clinical campuses. Information regarding the model, location,and dose detected from the device, as well as the treatment fields, fraction size, and treatment energy was collected.During this time, a new management policy for these patients had been implemented requiring treatment with low-energy beams (6 MV) and close surveillance of the patients in partnership with their electrophysiologist, as theyreceived RT.Results: During the study period, 33 patients were treated with an ICD in place. One patient experienced a defaultof the device to its initial factory setting that was detected by the patient hearing an auditory signal from the device.This patient had initially been treated with a 15-MV beam. After this episode, his treatment was replanned to becompleted with 6-MV photons, and he experienced no further events.Conclusion: Patients with ICDs and other implanted computer-controlled devices will be encountered more fre-quently in the RT department, and proper management is important. We present a policy for the safe treatmentof these patients in the radiation oncology environment. � 2009 Elsevier Inc.

Implanted cardiac defibrillator, Radiotherapy, Patient monitoring.

INTRODUCTION

Recent advances in the field of cardiology have improved the

longevity of patients with acute myocardial infarction and

heart failure. These developments have come from both the

medical and technical arenas, with two trials demonstrating

an advantage in the prophylactic use of implantable cardiac

defibrillators (ICDs). These two studies, the Sudden Cardiac

Death in Heart Failure trial (1) and the Defibrillators in Non-

ischemic Cardiomyopathy Treatment Evaluation (2), demon-

strated a 27% and 23% decrease in sudden death,

respectively, with the prophylactic use of ICDs compared

with medical management alone. The cost-effectiveness of

implanting and maintaining the devices compared with

chronic medical treatment of these patients has also been es-

tablished (3, 4). As a result, Medicare and Medicaid Services

have agreed to expand their coverage of these devices to in-

clude the preventative setting. During the past 10–20 years,

the devices have become smaller, and the implantation proce-

dure has become much less morbid. These factors will

presumably translate into an increased likelihood that these

152

patients will thrive with their cardiac conditions and require

radiotherapy (RT) for a cancer diagnosis in their lifetime.

The proper treatment of these patients in the RT depart-

ment in the era of high-energy linear accelerators is not

well understood. The American Association of Physicists

in Medicine last proposed guidelines for the treatment of pa-

tients with implanted cardiac devices in 1994 (5). This docu-

ment concentrated on the treatment of patients with

implanted cardiac pacemakers and only commented on the

emergence of implantable defibrillators and the need to re-

view their function in the future. Other groups in Europe

have also begun to explore the topic (6, 7) and have suggested

from in vivo testing that some signal detection interference

might occur when the device is placed in the radiation field.

The dose tolerance of the devices has been studied (6, 8, 9),

and it has been shown that the complementary metal oxide

semiconductor used is the most dose-sensitive component.

Within this, the random access memory (RAM) holds the in-

dividual patient pacing and detection threshold information.

To protect this element, it has been proposed that the total

dose seen by the defibrillator should not exceed 2–5 Gy,

Reprint requests to: Daphna Y. Gelblum, M.D., Department ofRadiation Oncology, Memorial Sloan-Kettering Cancer Center,650 Commack Rd., Commack, NY 11725. Tel: (631) 623-4200;Fax: (631) 864-5786; E-mail: dgelblum@mskcc.org

Conflict of interest: none.Received Jan 25, 2008, and in revised form June 19, 2008.

Accepted for publication June 20, 2008.

5

1526 I. J. Radiation Oncology d Biology d Physics Volume 73, Number 5, 2009

depending on the report. Company literature reflects these

findings as well (10–12), and some manufacturers have

been unwilling to quote a minimally acceptable dose.

Three potential interactions can occur between the ICD

and radiation generation and therapy delivery. The first is re-

lated to the radiation dose delivered to the device. The second

is the unintentional firing of the device during treatment as

a result of radiofrequency interference from the linear accel-

erator. Third, and perhaps most important, is the possible re-

programming of the device from some poorly understood

interference from the linear accelerator or the radiation

beam itself. The problem of reprogramming the defibrillators

is a stochastic event that we have now encountered on two

occasions in 2 separate patients during the past 3 years.

At Memorial Sloan-Kettering Cancer Center, we are en-

countering greater numbers of these patients and as a result

have looked to the manufacturers’ data and company techni-

cal support more closely for guidelines on how to handle

these patients safely. We have found a broad spectrum of

opinion with few supportive data. One school of thought is

that no interference has been documented between the

high-energy linear accelerators used in oncology and the

function of the defibrillator devices. Thus, beyond a discus-

sion with the treating cardiologist, no further special monitor-

ing has been advised, aside from minimizing the dose to the

device and interrogating it before the start, and after the com-

pletion, of the RT course. On the more conservative side,

other sources have suggested daily suspension of tachycardia

detection and therapy delivery by magnet placement during

RT sessions and frequent device interrogation to confirm

continued functioning. Published cardiac device malfunction

registries only track devices that require explantation and re-

placement, which is typically not required in the radiation set-

ting; thus, they do not capture the events of concern in our

patient population (13, 14). A survey of current practice in

the United States was undertaken by Solan et al. in 2004

(15). Their review underscored the lack of understanding

that exists in the clinical office practice of radiation oncology

and has prompted us to study this issue further.

METHODS AND MATERIALS

The sentinel case in our department that prompted our policy re-

vision was witnessed in a patient found to have a malfunctioning

ICD who had been treated for prostate cancer. As was our standard

policy at the time for all patients with implanted cardiac devices, a di-

ode was placed on his defibrillator (which was positioned outside of

the radiation portal in his left chest) on the first day of his RT ses-

sions to document the dose delivered to the device on a daily basis.

Additionally, a rhythm strip was run to confirm normal cardiac

rhythm during RT. He was treated with 15-MV photons and re-

ceived a total dose of 8,640 cGy in 48 fractions uneventfully. On

a follow-up visit with his cardiologist, it was found that his cardio-

defibrillator had reverted back to its initial factory programmed set-

tings. From the ICD’s memory, it was established that the

reprogramming had occurred immediately after one of his daily

RT visits. The device was reprogrammed and interrogated, with

the finding that the patient had not experienced any cardiac events

during that time, and he has not experienced any subsequent events.

In an attempt to quantify the incidence of these devices in our pa-

tient population and to capture any RT-associated malfunctions, a re-

view of our patients was conducted. After obtaining permission

from the institutional review board, all physicians in our main facil-

ity in Manhattan, as well as in our four regional offices, were asked

to complete a brief questionnaire documenting the specifics of the

case and submit it to a central data collection site when they encoun-

tered and treated patients with ICDs. This report presents our expe-

rience between June 2005 and December 2007. The data collection

began after an ad hoc committee, composed of radiation oncologists,

radiation nurses, cardiologists, and physicists, had already estab-

lished new patient management guidelines in light of the recent ex-

periences in the department (Fig. 1). It was decided that every

patient’s case would be individually discussed with their electro-

physiologist and that, at a minimum, no patient would be treated

with an energy >6 MV to prevent stray neutron emission, a suspected

cause of electrical interference. On the first day of treatment, a ther-

moluminescent dosimeter would be placed on the device to docu-

ment the dose delivered to the device, and a rhythm strip would

also be run during the entire first treatment delivery to visualize

the cardiac rhythm. Additional manipulation and monitoring of

the device was then individualized according to the specific manu-

facturer’s recommendations and the recommendations of the pa-

tient’s private electrophysiologist. This additional monitoring

ranged from daily post-treatment device interrogation in patients

deemed to be device dependent and who had a history of proper de-

vice firing in the past, to educating patients with self-check devices

to use the magnet to verify proper ICD functioning themselves, to

weekly interrogation by a company representative.

RESULTS

During the study period, 12,972 patients were treated at the

radiation oncology department of Memorial Sloan-Kettering

Cancer Center. Of these patients, 33 had ICDs (Table 1). Of

the 33 patients, 22 had been treated with definitive intent and

11 palliatively. Two patients required relocation of the device

out of the proposed radiation portal before starting RT. Three

patients were treated sequentially to two separate fields for

multifocal disease. Two patients had their prescribed treat-

ment dose limited because of dose constraints posed by the

presence of the defibrillator. The total dose to the ICDs was

1–299 cGy (Table 1). The patient management policy

followed is outlined in Fig. 1.

During the study period, 1 additional patient experienced

a resetting of his device during therapy (Patient 5, Table 1).

This patient was receiving treatment of rectal cancer and

again had been treated with 15-MV photons. He had a Med-

tronic device, which is built to send an audible alert when the

device reverts to the factory settings. This occurred after only

four treatments, although the patient stated that he was not

aware of the tone for an additional 5 days. The device had

been located well out of the radiation portal in the left chest

and had been interrogated before and after his first RT session

without incident. After consultation with this patient’s elec-

trophysiologist, the company representative was called,

who reprogrammed the device. He completed therapy with

6-MV photons. For the remainder of the RT, a rhythm strip

Defibrillator care in radiation d D. Y. GELBLUM AND H. AMOLS 1527

Fig. 1. Newly adopted patient management guidelines for patients with implantable cardiac defibrillators who are under-going treatment in the Department of Radiation Oncology at Memorial Sloan-Kettering Cancer Center (New York, NY).

was run and observed daily at the treatment machine, and the

company representative was on site for daily, post-RT inter-

rogation of the defibrillator. He completed his cancer therapy

successfully to a total dose of 50.4 Gy in 28 fractions and has

had no further issues. This patient was treated at the time that

the institutional policy was being formulated and was ini-

tially treated with 15-MV beams. Since then, we have treated

all patients with ICDs with low-energy 6-MV beams and

have not had any further events. We are confident with the

treatment plan defined.

The ICD in 3 patients was exposed to >2 Gy without expe-

riencing ICD malfunction, supporting our hypothesis that

this is not a dose-driven effect.

DISCUSSION

As implantable medical devices become more common

and successful in permitting the population to age, the fre-

quency of these encounters in RT departments will increase.

It is incumbent on the radiation oncologist to understand the

interaction of our linear accelerators with these devices to en-

sure the safety of our patients. The total dose delivered to the

device is known to be a hazard to the function of any mechan-

ical device located in or near a radiation field; thus, all im-

planted devices (e.g., pacemakers, infusion pumps, ICDs)

should be kept outside the primary radiation beam. Clinicians

and scientists unfamiliar with therapeutic radiation have ad-

vised ‘‘shielding the device’’ if it is in the direct beam. In

the case of a typical 6-MV beam, >5 cm of lead would be nec-

essary to attenuate the beam by 90%; thus, a simple lead

apron would not be effective and might even increase the po-

tential for scatter dose. With the increased use of intensity-

modulated RT, an often forgotten introduction of increased

photon scatter propagated from the head of the linear acceler-

ator has occurred. More conformal treatment delivery has

also permitted dose escalation, which also leads to increased

scatter. Thus, the only solution is to physically separate the

device from the radiation portal, in some cases requiring sur-

gical relocation of the device.

The issue of possible aberrant firing of the defibrillator

is of most concern with St. Jude’s medical products (11).

This seems to be a hypothetical occurrence that, according

to all three of the leading manufacturers in the United

States, has never been known to have occurred. The con-

cern is that the radiofrequency generated at the head of the

linear accelerator as the photon beam is generated would

cause the ICD to inappropriately deliver therapy in the ab-

sence of cardiac arrhythmia. Because other environmental

settings exist in which ICDs can misfire, we have thought

that protecting against this possibility might not be war-

ranted, although educating the patient and staff about

this possibility is.

1528 I. J. Radiation Oncology d Biology d Physics Volume 73, Number 5, 2009

Table 1. Treatment parameters for 33 patients with implanted cardiac defibrillators

Pt.No. Manufacturer

Treatmentfield

Defibrillatorlocation

Energy(MV*)

Totaldose(cGy)

Dose toAICD(cGy*) Comment

1 St Jude Scalp Left abdomen 6 MeV 6,600 6.62 Medtronic Right lung mass Left superior chest 6 8,000 163 Medtronic Left flank mass Left superior chest 6 3,000 144 Medtronic Whole brain Left superior chest 6 3,000 205 Medtronic Pelvis Left chest wall 6/15 5,040 206 St. Jude Left hip Left superior chest 6 3,000 47 St. Jude Whole brain Right superior chest 6 3,000 468 St. Jude Left hip Left chest 6 3,000 49 Guidant Lung Left abdomen 6 6,480 18

10 Guidant Left lung apex Left abdomen 6 7,560 1811 Medtronic Prostate Left superior chest 6 5,040 21 Also had HDR implant12 Guidant Prostate Left superior chest 6 8,100 14 Gy13 Contak Prostate Left superior chest 6 8,100 22.514 St Jude Right hip Right chest wall 6 2,400 815 St Jude Right parietal brain Left anterior chest wall 6 1,335 1316 Medtronic Brain Left chest wall 6 5,940 29.717 ? Right lung/bilateral

mediastinumLeft chest wall 6 4,600 246.1 Treatment dose limited

by dose todefibrillator

17a Brain Left chest wall 6 Second treatment course18 Medtronic Prostate ? 6 8,100 2719 Medtronic Prostate Left chest wall 6 8,100 13.520 Medtronic LUL, bilateral

mediastinumand rightsupraclavicular

Left chest wall 6 4,600 299 Treatment limitedbecause unable tomove defibrillator toacceptable place

21 Guidant C4–T3 andsubsternalnodule

Right upper chest wall 6 MVand 9MeV

600 6.6 Treatment stoppedbecause patientdeclining clinically

22 St Jude Pelvis Left chest wall 6 3,420 9.5 Treatment stoppedbecause of noncardiaccomorbidities

23 Medtronic Left parietal brain Left chest wall 6 4,005 3024 Medtronic Prostate and nodes Left chest wall 6 8,100 2725 Medtronic Right distal

femur and leftproximal femur

Left upper chest wall 6 2,000 and 2,000 1.0

25a Medtronic Sternum Left upper chest wall 6 2,000 8 Second sequentialtreatment course

26 Guidant Whole brain Left abdomen 6 3,000 1027 Guidant Right chest wall Left shoulder 6 5,000 117.528 Medtronic Trachea and right

supraclavicularLeft clavicle 6 6,000 210

29 Guidant Left parietal scalp Left abdomen 6 3,600 7.230 Medtronic Right groin Left chest wall 6 3,600 431 Medtronic Right upper lobe Left clavicle 6 4,800 119.6 Hypofractionated31a Left lower lobe Left clavicle 6 4,800 32.4 Hypofractionated;

second sequentialtreatment; severalweeks after treatment,patient hospitalizedwith arrhythmia

32 Guidant Left upper lobe Left abdomen 6 7,560 21 Defibrillator movedfrom left upper chestto abdomen to permitRT

33 Medtronic Right neck Left chest wall 6 3,000 23

Abbreviations: Pt. No. = patient number; AICD = automatic implantable cardiac defibrillator; HDR = high-dose-rate; LUL = left upper lung;RT = radiotherapy.

* Unless otherwise noted.

Defibrillator care in radiation d D. Y. GELBLUM AND H. AMOLS 1529

Few reports have been published documenting the possi-

ble interactions between RT and ICDs, although a few case

reports have been submitted. Thomas et al. (16) reported in

a cardiac journal, a case report of a patient with a history of

coronary artery disease, myocardial infarction, coronary ar-

tery bypass, hypertension, atrial fibrillation, and right upper

lung cancer. The patient had had an ICD implanted in the

left chest wall. After 56 Gy delivered to his right-sided

lung cancer in 2.0-Gy/d fractions with 18-MV photons, he

was seen by his cardiologist. On interrogation of the ICD,

it was found to have reset to its fallback setting 9 days earlier

during RT session (determined by comparing the date and

time of the stored event and the patient’s therapy schedule).

No other malfunction was found in the device after additional

testing. They speculated that not only is direct or scattered

dose to the device of concern, but also the potential for elec-

tromagnetic interference of the ICD from the linear accelera-

tor. Hoecht et al. (17) in a letter to the editor briefly

mentioned an ICD malfunction that had occurred in a patient

undergoing RT in their department. The exact details of the

treatment parameters were not discussed, but they did men-

tion that the ICD was seated in a pectoral pocket, and the

RT was delivered to the patient’s pelvis. The malfunctioning

device was found to have gone into a fallback programming

to the fixed detection area and a fixed pace rate of 50/min.

Nemec (18) published a case report of a patient treated to

a thoracic tumor in the left chest with a Guidant ICD also

seated in the left chest, as documented by an accompanying

X-ray film. The radiation energy used for treatment was not

documented, and mention was made that the treating physi-

cians did not find it necessary to relocate the ICD before treat-

ment. During the third session (180-cGy fractions), the

patient collapsed and was found to have wide complex tachy-

cardia. The patient recovered, and the device interrogation

after explantation because of feared radiation damage failed

to show any malfunction. The manufacturer suspected a scat-

tered radiation effect on the RAM. The total radiation dose to

the device was not known.

Sepe et al. (19) reported on their experience treating a pa-

tient for larynx cancer with 6-MV photons to a total dose of

60 Gy. During the course of therapy, the ICD received a cu-

mulative dose of 2.5 Gy. They continually monitored the

device throughout the therapy and immediately after comple-

tion of the RT course and found no adverse effects. They sug-

gested a similar management policy as put forth in the present

report, with the exception of a blanket recommendation to

disarm the ICD on a daily basis for treatment, with continual

electrocardiographic monitoring.

It is not only the direct photon exposure of a device that

must be considered, but also the potential for a single hit neu-

tron particle interaction. Much like radiation damage to a liv-

ing cell, radiation damage to ICDs and similar solid state

devices can, theoretically, manifest by way of two physically

different pathways, the so-called somatic or stochastic path-

ways. Somatic effects, such as skin erythema or loss of infor-

mation stored in the ICD’s memory, can result from

accumulation of a low-energy transfer (LET) dose, for which

no threshold dose has been established and for which the

magnitude of the effect is proportional to the absorbed

dose. Catastrophic or stochastic events, such as cell mutation

or total ICD failure, however, can also result from the passage

of a few, or even a single high LET particle through a cell nu-

cleus or an ICD chip. Stochastic effects are usually all or

nothing bimodal events—either the catastrophic happens or

it does not. Another example of a random catastrophic ICD

effect would be emitting a high-voltage shock without detec-

tion of cardiac arrhythmia. Unlike somatic effects, the prob-

ability of occurrence of a stochastic event is dose dependent,

but the magnitude of the damage or failure is not.

Published data have documented that relatively high radi-

ation doses (>1 Gy) do indeed cause cumulative damage to

ICDs, and virtually every manufacturer has recommended

that ICDs not be placed directly in the radiation beam, even

if it requires surgical repositioning. When such precautions

are taken, none of our findings nor any published reports

that we are aware of have suggested that small (<1 Gy) radi-

ation doses should be of concern to RT patients with ICDs.

Similarly, we know of no documented cases of ICD failure

resulting from the electromagnetic fields produced by high-

energy linear accelerators. Nonetheless, scattered cases

have been reported, including our own, of ICD failure, mem-

ory loss, reprogramming, and so forth, occurring during or

shortly after a RT session. In the events reported in the pres-

ent study, the failure was a reprogramming error (the ICD re-

programmed to its default values) occurring sometime during

the RT course.

Bradley and Normand (20) have reported on extensive

measurements of so-called single event upsets (i.e., sto-

chastic events) in ICDs. Because of the paucity of data

for RT patients, the study by Bradley and Normand was

based entirely on ICD failures observed in the general pop-

ulation that were caused, presumably, by background radi-

ation, plus direct irradiation of ex vivo ICDs under

controlled laboratory conditions. Their data suggested

that the RAM is the subsystem in the ICD most sensitive

to radiation damage because of the small amount of charge

stored in the RAM and its resulting high volatility. Perhaps

more significantly, they reported that the cross-section of

probability of single event upsets is strongly dependent

on LET, increasing by approximately three orders of mag-

nitude as the LET increases from 2 MeV-cm2/mg to 15

MeV-cm2/mg. X-rays and electrons have LETs of about

2 MeV-cm2/mg, and neutrons have LETs of approximately

5–10 MeV-cm2/mg. Although highly speculative, this is

consistent with ICD events observed in vivo in RT patients

only for the high-energy 15-MV X-ray beam for which

neutron contamination is about 0.1% of the total dose.

We speculate that the ICD events observed in RT patients

who have received relatively low radiation doses might be

caused by neutrons rather than by X-rays or other electro-

magnetic radiation. The total dose seen by the defibrillator

in the 2 patients in our study who experienced reprogram-

ming was low, and those few patients in whom the ICD

was exposed to >2 Gy had no issues.

1530 I. J. Radiation Oncology d Biology d Physics Volume 73, Number 5, 2009

The whole body effects of neutron contamination from

high-energy photon beams have been shown by Hall et al.(21). In their work, they demonstrated that when a phantom

was radiated to the pelvis in fields similar to those used for

standard four-field treatment of cervical cancer, the dose re-

ceived by scatter to the breasts was more significant when

the treatment was administered with 20-MV than with

6-MV beams. The dose seen by thermoluminescent dosime-

ter measurements placed in tissue at the level of the breast

was 0.26 Gy for 6-MV and 0.55 Gy for 20-MV treatment

when 70.2 Gy was delivered in a fractionated manner to

the pelvis. A bubble meter was used to quantitate the neutron

dose and detected a 0.005-Gy neutron dose to the tissue flask

in the breast location. This correlated with the calculated ex-

pected maximal neutron dose of 0.015 Gy, assuming a maxi-

mal neutron production rate of 3� 1012 neutron/Gy of X-rays

delivered. This form of radiation delivery to implanted de-

vices cannot be protected against by external blocking mech-

anisms and can only addressed by altering the treatment

delivery to minimize the chance of these interactions occur-

ring. We, therefore, advocate that patients with ICDs be

treated with low energy (<10-MV) photons whenever possi-

ble. Since the institution of that policy, we have not detected

any further reprogramming events. We also continue to vig-

ilantly observe these patients, together with their cardiolo-

gists, as they go through treatment.

The sensitivity of modern devices to RT has not been well

documented. The three leading American manufactures have

failed to establish a ‘‘safe’’ or acceptable dose to the ICD, ac-

knowledging that the remaining issue is scattered dose,

which is difficult to quantitate. To date, the only group to

have studied and published on dose tolerances is Uiterwaal

et al. (22) from The Netherlands. They reported their obser-

vations after irradiating 11 devices from four manufacturers

(the three major American companies and Biotronik, which

is European based). They used 6-MV photon beams to a cu-

mulative dose of 20 Gy. Of the 11 devices they observed, 4

devices had ‘‘complete loss of function’’ after only 1.5 Gy.

They concluded that these are devices are very radiosensitive

and that new guidelines for the treatment of these patients are

indicated. As we await these parameters, we have been suc-

cessful in implementing the policy put forth in the present

report.

CONCLUSION

The complexity of patient treatment in the radiation oncol-

ogy department continues to increase as technology prog-

resses. As evidenced in the present report, the advances

made in other fields of medicine can directly effect how we

function in our departments. The hypothesis that ICD dam-

age is caused by high LET neutron scatter has been proposed

by several investigators, and the clinical evidence presented

in our report supports this theory. We will continue to watch

our patients carefully and urge other centers to do so as well.

When a patient event is witnessed in which the ICD was not

exposed to excessive radiation as a result of therapy and the

ICD malfunctions, this should be reported.

We have adopted and advocate a patient management pol-

icy for patients with ICDs who require RT. We suggest treat-

ment with low-energy beams (<10 MV), keeping the ICD as

far from the treatment field as possible, discussing each indi-

vidual patient’s care with their cardiologist, and determining

the optimal practice. Interrogation should always occur be-

fore and after the initial fraction and again with some fre-

quency thereafter, as determined in consultation with the

patient’s electrophysiologist. A thorough discussion must

be held with the patient to educate them about the issue as

it applies to them and their specific device. The 2006 Journal

of the American Medical Association cardiac patient educa-

tion sheet does not mention RT in their section on ‘‘Concerns

for patients with an ICD’’ (23). It might also be determined

that a patient is so dependent on the defibrillator that they

should not be considered for RT.

The treatment of these patients using proton machines

requires a separate review.

REFERENCES

1. Bardy GH, Lee KL, Mark DD, et al. Amiodarone or and im-plantable cardioverter-defibrillator for congestive heart failure.N Engl J Med 2005;352:225–237.

2. Sanders GD, Hlatky MA, Owens DK. Cost-effectiveness of im-plantable cardioverter-defibrillators. N Engl J Med 2005;353:1471–1480.

3. Glodman L. Cost-effectiveness in a flat world—Can ICDs helpthe United States get rhythm? N Engl J Med 2005;353:1513–1515.

4. Kadish A, Dyer A, Daubert JP, et al. Prophylactic defibrillatorimplantation in patients with nonischemic dilated cardiomyop-athy. N Engl J Med 2004;350:2151–2158.

5. Marbach JR, Sontag MR, Van Dyk J, et al. for the American As-sociation of Physicists in Medicine. Management of radiationoncology patients with implanted cardiac pacemakers: Reportof AAPM Task Group No. 34. Med Phys 1994;21:85–90.

6. Uiterwaal GJ, Springorum BGF, Scheepers E, et al. Influence oftherapeutic irradiation on the latest generation of implantablecardioverters/defibrillators. Europace Supplements 2004;6:96.

7. Uieterwaal GJ, Hurkmans CW, Springorum BF, et al. Interfer-ence signals detection by implantable defibrillators induced bytherapeutic radiation therapy [Abstract]. Heart Rhythm 2004;1:S78.

8. Niehaus M, Tebbenjohanns J. Electromagnetic interference inpatients with implanted pacemakers or cardioverter-defibrilla-tors. Heart 2001;86:246–248.

9. Hurkmans CW, Scheepers E, Springorum BG, et al. Influenceof radiotherapy on the latest generation of implantable cardi-overter-defibrillators. Int J Radiat Oncol Biol Phys 2005;63:282–289.

10. Medtronic CRM Technical Services. Medtronic company liter-ature on radiation tolerance of their implantable cardiac pace-makers and implantable cardiac defibrillators. Minneapolis,MN: Medtronic; 2006.

11. St. Jude Medical Cardiac Rhythm Management Division.St. Jude company literature on radiation tolerance of theirimplantable cardiac pacemakers and implantable cardiac defi-brillators. Austin, TX: St. Jude; 2006.

Defibrillator care in radiation d D. Y. GELBLUM AND H. AMOLS 1531

12. Guidant Corporation, Cardiac Rhythm Management Technical

Services. Guidant company literature on radiation tolerance of

their implantable cardiac pacemakers and implantable cardiac

defibrillators. Indianapolis, IN: Guidant Corporation; 2006.13. Maisel WH, Moynahan M, Zuckerman BD, et al. Pacemaker

and ICD generator malfunctions: Analysis of Food and Drug

Administration reports. JAMA 2006;295:1901–1906.14. Maisel WH. Pacemaker and ICD generator reliability: Meta-

analysis of device registries. JAMA 2006;295:1929–1934.15. Solan AN, Solan MJ, Bednarz G, et al. Treatment of patients

with cardiac pacemakers and implantable cardioverter-defibril-

lators during radiotherapy. Int J Radiat Oncol Biol Phys 2004;

59:897–904.16. Thomas D, Becker R, Katus HA, et al. Radiation therapy-in-

duced electrical reset of an implantable cardioverter defibrillator

device located outside the irradiation field. J Electrocardiol

2004;37:73–74.

17. Hoecht S, Rosenthal P, Sancar D, et al. Implantable cardiac de-fibrillators may be damaged by radiation therapy. J Clin Oncol2002;20:2212–2213.

18. Nemec J. Runaway implantable defibrillator—A rare complica-tion of radiation therapy. Pace 2007;30:716–718.

19. Sepe S, Schaffer P, Krimmel K, et al. Irradiation treatment oflaryngeal cancer in a patient with an implantable cardioverter-defibrillator (ICD). Onkologie 2007;30:378–380.

20. Bradley PD, Normand E. Single event upset in implantable car-dioverter defibrillators. IEEE Trans Nuclear Sci 1998; 452929–2940.

21. Hall EJ, Martin SG, Amols H, et al. Photoneutrons from med-ical linear accelerators—Radiobiological measurements andrisk estimates. Int J Radiat Oncol Biol Phys 1995;33:225–230.

22. Uiterewaal GJ, Springorum BGF, Scheepers E, et al. Influenceof therapeutic irradiation on the latest generation of implantablecardioverter/defibrillators. Eurospace 2004;6(Suppl.):96.

23. JAMA Patient Page: Implantable cardioverter-defibrillators.JAMA 2006;295:1964.