To define the practical clinical guidelines that can be implemented by busy radiation oncology departments to minimize the risk of harm to patients with implanted cardiac pacemaker (ICP) and implantable cardioverter-defibrillator (ICD) devices during radiotherapy.  Although the risk of potentially life-threatening malfunction secondary to electromagnetic interference (EMI) or ionizing radiation (IR) is recognized, clinical practice guidelines for the management of patients with modern ICP and ICD devices are occasionally inconsistent and inexplicit.

The usual indication for placement of an ICP is the prevention of bradycardia and its resultant symptoms of syncope, near syncope, and, occasionally, cardiac arrest caused by sinus node dysfunction or atrioventricular block. A more recent indication includes biventricular pacing in the treatment of congestive heart failure. Pacemakers are placed for both permanent and intermittent bradycardia. Atrioventricular block is frequently permanent and more often associated with severe and potentially life-threatening bradycardia.

Pacemaker systems consist of a pulse generator, containing a battery and electronics, connected to transvenous pacing leads in the right ventricle and sometimes in the right atrium and left ventricle, generally placed in a subcutaneous and infraclavicular location. Most pacemakers in the United States have leads in the right atrium and ventricle. The pacemaker may sense electrical activity in the chamber and be inhibited. If the heart rate falls below a programmed rate, it sends out an electrical current that should “capture” the paced chamber, resulting in an action potential being propagated through the chamber and causing mechanical contraction.

Implantable cardioverter-defribillators are implanted to treat symptomatic ventricular tachycardia and fibrillation or as prophylaxis to prevent sudden cardiac death in high-risk patients with new, more liberal indications. Modern systems consist of a generator containing a battery, capacitor, and electronics connected to a single right ventricular defibrillator lead or, sometimes, with a right atrial or left ventricular lead. All ICDs have pacemaker capability and antitachycardia pacing algorithms.

Pacemaker batteries are monitored for signs of depletion by placing a magnet over the pulse generator. This closes a magnetic reed relay switch converting the pacemaker to a fixed rate with no sensing function. When the battery is depleted to a certain point, the rate, and sometimes the mode, will change to indicate that it is time for elective replacement. Magnets placed over ICDs usually inhibit antitachycardia pacing and shocking functions without affecting their pacemaker function. All functions return to baseline when the magnet is removed.

A typical “demand” pacemaker works by monitoring the heart rate and firing at a fixed rate only when bradycardia develops. Similar to the relay switch closure induced by placing a magnet over the pacemaker to evaluate battery depletion, ICP exposure to electromagnetic fields produced by linear accelerators and betatrons may result in loss of pacemaker inhibitory function and inappropriate firing.

Pacemaker functioning can be adjusted by sensing of electromagnetic energy emitted by programmers. Similarly, ICPs can be affected by external electromagnetic fields. This may result in sensing the field as a myocardial potential, thus inhibiting its output. Other malfunctions may include closing the reed switch, resulting in fixed rate pacing, triggering of output, or more serious functional impairment, including device reprogramming. Fortunately, most pacemaker interactions are reversible when the field exposure is eliminated.

In contrast, ICP damage caused by IR is often permanent and cumulative and eventually will lead to failure of the device. Older pacemakers used nonprogrammable bipolar semiconductors for their circuitry, which proved to be fairly resistant to the effects of IR even when directly exposed to cumulative doses far in excess of those used in clinical practice. Advances in pacemaker technology led to the use of complementary metal oxide semiconductor (CMOS) circuitry in modern pacemakers. CMOS circuits are smaller, more reliable, and energy efficient but also more radiosensitive.

In a MOS transistor, gate, source, and drain terminals are composed of silicon, and silicone dioxide serves as an insulator between these various components of the circuitry. The voltage applied to the gate modifies the amount of current flowing in the channel between source and drain. When a semiconductor is exposed to high-energy-charged particles (IR), atoms become excited, resulting in electrons in the conduction band and holes in the valence band. Normally, after removal of the radiation source, these electrons and holes recombine and their energy is dissipated as heat. When MOS transistors are exposed to IR, electrons and holes are created within both the Si and the SiO2 insulator. However, rather than recombining as within the Si, the electrons within the oxide move in response to an electric field and leave the device. Holes in the valence band of the oxide are much less mobile and stay in place, leading to a net positive charge.

A build-up of positive charge in the SiO2 separating the gate and channel may alter the changes in current flow experienced in response to changes in the voltage applied to the gate. A build-up of positive charge in the SiO2 insulation between various components of the MOS circuit may allow formation of leakage currents within the circuit (6, 12). Because these aberrant pathways can occur in multiple locations and do so at random, the mode of pacemaker malfunction secondary to IR exposure is unpredictable. In general, malfunctions may include altered sensitivity, increased or decreased pulse width and frequency, or complete cessation of pacemaker function.

That radiation exposure may cause significant damage to ICPs and ICDs with potentially life-threatening results is well established. The cause of this damage may be either the strong electromagnetic fields produced by linear accelerators and betatrons (EMI) or IR itself. A recently published editorial summarizes effects from electronic metal detector gates, mobile telephones, electronic article surveillance systems, MRI, and electrocautery devices in addition to RT with proposed precautionary recommendations. The purpose of this study was to review the available clinical literature on the effects of RT on ICP and ICD devices, assess and compare manufacturer guidelines and clinical practice patterns regarding patient management during RT, and propose clinically pertinent policies and procedures that can be adopted into general practice.

Discussion 

Marbach and associates  were among the first to identify problems with CMOS devices due to RT. The authors exposed several pacemakers to radiation from both linear accelerators and betatrons under conditions simulating RT and encountered pacemaker failure or malfunction, most significantly attributed to EMI. They concluded that patients with cardiac pacemakers should not be treated with betatrons and recommended ECG monitoring during treatment with linear accelerators.

Adamec  went on to distinguish between radiation effects on demand vs. programmable pacemakers, demonstrating little significant effect of doses of 70 Gy to the demand units but “complete sudden failure” of programmable units at varying doses. Other authors have documented life-threatening pacemaker failure at radiation doses of <70 Gy and well within the therapeutic range that would be encountered in clinical practice. Souliman tested 18 previously untested CMOS circuitry pacemakers from five different manufacturers. Although EMI had no effect in this study, 11 of the 18 pacemakers failed secondary to IR at cumulative doses between 16.8 and 70 Gy and did not recover. Loss of inhibition manifested by fixed rate pacing in the “interference mode” preceded complete failure. Other effects noted were increased pulse interval and decreased pulse rate, similar to those observed from battery depletion. Although recovery of pacemaker malfunction may occur, this may be delayed until after RT completion and is usually incomplete, resulting in an unreliable device. Their observations prompted these authors to recommend patient monitoring both during and for several weeks after RT completion.

Recently, Mouton reported on the effect of both the cumulative radiation dose and the dose rate on 96 ICPs from various manufacturers, including the Guidant and Medtronic pacemakers available in the United States. The authors categorized the malfunctions according to the type and risk to the patient. Malfunctions such as temporary changes in pulse frequency and amplitude could be transient but still detrimental to the patient. Potentially life-threatening malfunctions included multiple pacemaker silences >5 s in duration seen in 6 pacemakers (6%) tested at cumulative doses <2 Gy. The number of pacemakers exhibiting potential life-threatening failures increased to 14 (15%) at a cumulative dose of 5 Gy. These collective observations have led numerous authors to recommend that direct exposure of the ICP to radiation should be avoided and that the device should be moved outside the field of radiation before the initiation of therapy

Implantable cardioverter-defribillators also use CMOS circuitry and are sensitive to both EMI and IR damage. The possible effects on ICDs from EMI include deactivation of shock therapy by reversion to the “off” mode because of closure of the reed switch, inhibition of brady pacing, inappropriate shocking caused by EMI being sensed by the ICD as ventricular fibrillation, interference with the ICD's ability to detect intrinsic cardiac activity and thereby causing asynchronous pacing at the programmed rate (noise response), and prevention of ICD interrogation and programming. Rodriguez exposed four ICD devices to 6-MV photons and 18-MeV electron radiation. Damage manifested primarily as a rapid increase in the detection and charging time with an increasing radiation dose and causing a decrease in battery capacity secondary to current drain. Niehaus and Tebbenjohanns found no evidence of damage, oversensing, or pacing inhibition at cumulative doses of <5 Gy.

Implanted cardiac pacemakers and ICDs may be affected by the electromagnetic fields generated by linear accelerators and betatrons, with fields from betatrons being stronger and more problematic. Interactions with pacemakers most commonly result in a transient “dropped beat” sensation of no clinical relevance that occurs most often as the treatment machine is turned on or off. Guidant and Medtronic guidelines suggest that painful shocking from ICD oversensing can be avoided by placing a magnet over the device during RT to temporarily suspend ventricular tachycardia and ventricular fibrillation therapies. EMI effects will reverse after the radiation equipment is turned off, and the magnet can be removed.

Ionizing radiation has a cumulative effect on ICDs that may manifest after completion of RT as altered memory. Guidant suggests this could result in limited programmability and loss of diagnostics so that the device invokes “safety mode,” or device failure, including the loss of shock therapy or pacing and loss of telemetry, diagnostics, and measurement/counter information. According to the Guidant recommendations, ICDs are more sensitive to radiation damage than ICPs by a factor of 5 to 10 times (18). The reason for this differential sensitivity offered by Guidant is that ICD operation instructions are stored in random access memory that can be damaged by scatter radiation. The company goes on to suggest that as ICP technology evolves, they may also become more sensitive to radiation exposure. Although Guidant does not suggest limits for either ICP or ICD devices, this information is consistent with Medtronic guidelines, suggesting lower cumulative dose recommendations for ICDs.

The first attempt at establishing management guidelines was published in 1989 in the informal setting of an American Society for Therapeutic Radiology and Oncology newsletter . These AAPM recommendations were updated in 1994 by the AAPM TG-34 but did not include recommendations specific to patients with ICDs. A specific practice policy recommendation initiated at the University of Michigan in response to these guidelines was published in 1991. The author offered a worksheet for documentation of policy implementation in the patient's treatment chart to ensure compliance.

Conclusion  

What appears to be needed is an updated set of guidelines that are readily implemented and universally applicable without regard to nuances or idiosyncratic differences between models and manufacturers, without undue reliance on other physicians for management decisions, and without undue burden to either the patient or department personnel. The conservative guidelines presented below are intended to streamline the process of delivering RT to patients with ICPs and ICDs by proposing “universal precautions.” They are a composite of recommendations from the published literature, device manufacturers, and practice guidelines already in place at some institutions. They are not meant to replace the existing AAPM TG-34 guidelines but to reinforce and expand on them and to suggest a practical means of ensuring their implementation. Salient differences between the current proposed guidelines and the 1989 AAPM and 1994 AAPM TG-34 guidelines are summarized below:

Radiotherapy can influence the functioning of pacemakers and implantable cardioverter-defibrillators (ICDs). ICDs offer the same functionality as pacemakers, but are also able to deliver a high-voltage shock to the heart if needed. Guidelines for radiotherapy treatment of patients with an implanted rhythm device have been published in 1994 by The American Association of Physicists in Medicine, and are based only on experience with pacemakers. Data on the influence of radiotherapy on ICDs are limited. The objective of our study is to determine the influence of radiotherapy on the latest generation of ICDs.

Results: Sensing interference by ionizing radiation on all ICDs has been demonstrated. For four ICDs, this would have caused the inappropriate delivery of a shock because of interference. At the end of the irradiation sessions, all devices had reached their point of failure. Complete loss of function was observed for four ICDs at dose levels between 0.5 Gy and 1.5 Gy.

Conclusions: The effect of radiation therapy on the newest generation of ICDs varies widely. If tachycardia monitoring and therapy are functional (programmed on) during irradiation, the ICD might inappropriately give antitachycardia therapy, often resulting in a shock. Although most ICDs did not fail below 80 Gy, some devices had already failed at doses below 1.5 Gy. Guidelines are formulated for the treatment of patients with an ICD. Introduction

With an aging population and the increased use of pacemakers and implantable cardioverter-defibrillators (ICDs), the incidence of patients presented for radiation therapy with a pacemaker or ICD is increasing. ICDs are pacemakers that are also able to deliver a high voltage shock in case of a life-threatening ventricular tachycardia or ventricular fibrillation.

The effects of radiation therapy on pacemakers has been studied in detail in the past. Based on the study results, the American Association of Physicists in Medicine (AAPM) has presented recommendations for irradiation of pacemaker patients in 1994. They used two categories to describe radiotherapy hazards. Electromagnetic noise interference (EMI) and radiation damage. They concluded that the EMI effects were transient and no study has reported serious problems with EMI using linear accelerators. The largest noise source had actually become the patient (i.e., muscle signals). They state that there is good evidence that transient interference from EM noise is not a problem around properly functioning contemporary radiation therapy equipment. Because the EM noise around linear accelerators has been even further reduced since the publication of this report and EMI effects are mostly transient in nature, EM effects are usually not considered to be a source of concern at present. However, the radiation dose to the patients is not decreased and even dose escalation studies have been implemented. Radiation damage effects are still a source of concern. The studies mentioned in the AAPM report encompass older pacemaker types based on old dipole and the first generation of complementary metal oxide semiconductor (CMOS) technology. Modern pacemakers, using newer CMOS circuitry, differ from these devices both in their sensitivity to radiation and type of malfunction observed. Since the AAPM report, only a few pacemaker studies have been performed

The effect of radiation therapy on ICDs has been studied even less. The first human ICD implant occurred in 1980, and Food and Drug Administration approval was granted in 1985. Although the number of implanted ICDs is not as high as the number of implanted pacemakers, their implantation rate is growing rapidly and exceeded 30,000 per year worldwide in 2001. Besides the effects that are also observed for pacemakers, ICD irradiation might also lead to a decrease in the ability to deliver the high-voltage shock. This malfunction might be due to failures in the ability to sense the heart signals that should lead to a shock delivery, a too-long period between the shock triggering and delivery, and a decreased or totally absent ability to deliver the shock. This can be potentially life-threatening for the patient.

Rodriguez et al. included 23 pacemakers and 4 ICDs in their study. The ICDs were irradiated with a 6-MV direct photon beam to a total dose of 250 Gy or failure. The ICDs were unable to deliver a shock because of sensing/detection problems (three devices at 51, 54, and 70 Gy) or the inability to charge (one device at 54 Gy).

Niehaus and Tebbenjohanns reported that no damage of the ICD, oversensing, or inhibition of pacing was observed for 3 patients who underwent radiation therapy with a cumulative dose of <5 Gy to the ICD.

Hoecht et al. reported ICD malfunction during the course of radiotherapy for 1 of 3 patients studied. After replacement, the new ICD suffered the same malfunction (fallback in fixed detection and pacing mode). Both devices received a dose of <0.5 Gy. Five additional devices were examined. At first, electromagnetic interference without irradiation was examined, then scatter radiation and at last direct exposure to the therapeutic beam. One ICD fell into fallback mode when exposed to scattered radiation, the other ICDs only exhibited malfunction after direct exposure of more than 50 Gy. All three devices that fell into fallback mode were of the same model and manufacturer.

Direct irradiation and fractionation

AAPM Task Group 34  recommends placing the pacemaker or ICD outside of the radiation beam. However, in our study, the devices were placed in the direct beam to limit the actual irradiation time. We assume that in accordance with Mouton et al., that the ionizing effects of scattered radiation outside of the unshielded beam, have the same effects at the same dose level on the pacemakers as direct irradiation. The possible mechanisms of damage by irradiation on pacemakers have been described in detail by Last.

Exposure to ionizing radiation leads to the formation of excess electron-hole pairs in both the silicon and the silicon dioxide insulator. The excess pairs rapidly recombine (in 1–10 ?s) when the radiation stops. With the beam setup used in our study (6 MV, 6 Gy/min at the ICD, pulse repetition frequency = 600, pulse duration = 3.2 × 10–6 s), the time between consecutive pulses = 1.66 × 10–3 s. The instantaneous dose rate is approximately 50 Gy/s. This implies that all excess pairs in the silicon created during one pulse have already long recombined before the next pulse is given. This holds for all beams of all medical linear accelerators that are used today. Transient effects caused by photocurrents generated by a high-dose rate begin to be appreciable for dose rates >104 Gy/s and thus are unlikely to be significant at the relatively low instantaneous dose rate of 50 Gy/s. The instantaneous dose rate of pulsed linear accelerators is much higher than the dose rate of cobalt-60 beams or kilovoltage X-ray beams. It is therefore reasonable to assume that these transient effects are also irrelevant in cobalt-60 beams or kilovoltage X-ray beams.

In oxide, the electrons soon leave by flow toward the metal or semiconductor. This might cause permanent and transient effects. The holes in the valence band are not very mobile and tend to be attracted by structural defects and remain there, resulting in a buildup of trapped positive charge in the oxide. This, by definition, is a cumulative, dose rate independent effect.

Thus, apart from transient effects during the irradiation itself, no large differences are to be expected in pacemaker response between direct and scattered irradiation at similar cumulative dose levels. As also observed in previous studies, some devices recover from irradiation induced malfunctioning. The time to recover can be in the order of hours or even weeks. Thus, besides the cumulative dose, the irradiation fractionation scheme used to irradiate a patient might also be of some influence on ICD functioning. Further study is needed to verify this assumption.

We assumed that a different beam energy or type (photons or electrons) would have the same effect on the ICDs. Only one large study has been published describing the irradiation of pacemakers and ICDs with different beam energies (5). Rodriguez et al. did not find any dependence using 6 MV photon beams (17 pacemakers and 4 ICDs) and 18 MeV electron beams (7 pacemakers).

Irradiations were performed at room temperature, similar to most previously reported large pacemaker and ICD studies (2, 5, 8, 11, 13). Adamec et al. performed irradiations at 37°C, but did not report any temperature-dependent effects (1). Venselaar et al. also performed irradiations at 37°C, and found that pacemakers that seemed to recover at 37°C, sometimes revealed defects at temperatures other than 37°C (3). Mouton et al. measured an increased battery impedance at 37°C compared with 23°C. However, they did not report any effect on pacemaker operation resulting from this difference. It seems reasonable to assume that irradiation effects at room temperature will closely resemble effects that could be observed at body temperature. We hypothesis that the irradiation fractionation schedule will have a larger influence on some (transient) radiation effects.

EM interference

As discussed in the Introduction, substantial evidence exists that electromagnetic noise interference in pacemakers and ICDs is negligible around modern linear accelerators. We refer to a review of Last for a thorough review of the data on this subject. The most commonly observed EM influence on pacemaker/ICD operation is a temporary pacing inhibition when the accelerator is switched on or off. For the three ICDs (4, 9, and 11) that exhibited inhibition during irradiation, we tested whether the inhibitions were caused by the EM field. After inhibition was observed during the irradiation up to 20 Gy, the ICDs were placed outside the primary beam, but with the leads outside and inside the beam, respectively. Both times, 6 Gy was given at the isocenter. No inhibitions were observed and no interference detection was observed. Hence, EM interference did not seem to be the cause of the inhibitions observed while the ICD devices were irradiated.

Recommendations

Implantable cardioverter-defibrillators of different manufacturers show a large variation in their sensitivity to radiation and there does not seem to be a way to predict the ICD sensitivity to irradiation of patients referred for radiotherapy. Moreover, modern ICDs seem to be more sensitive to radiation than pacemakers. If tachycardia monitoring and therapy are functional (programmed on) during irradiation the ICD might inappropriately give antitachycardia therapy, often resulting in defibrillation shock if erroneous ventricular fibrillation or ventricular tachycardia is detected. So far, no international guidelines for the treatment of ICD patients have been published. We recommend implementation of the following guidelines.

Determine, together with the other physicians involved in the oncology care in your hospital, which treatment is optimal for your patient. Take into consideration the possible effects radiotherapy may have on ICDs, as well as the possible increased risks from, for example, surgery or chemotherapy from the patients’ heart condition.

When radiotherapy is part of the treatment:

The ICD should always be located outside the irradiation field.

The absorbed dose to be received by the ICD should be estimated before treatment for documentation purposes. Estimation methods can be found in the literature

Program the ICD temporarily to “monitor only” before each individual irradiation fraction. After consultation with the patient’s cardiologist, consider switching the ICD to “monitor only” before the first irradiation fraction and only switch back to therapy mode after the complete treatment is given. Consider that even if the ICD is turned off and on with every treatment fraction, no guarantee can be given that the ICD is still able to deliver a shock if needed.

Monitor the ECG and have ICD-qualified personnel stand by at every fraction. The treating radiation oncologist might consider omitting (part of) these safety measures if consensus is reached on this aspect with the patient’s cardiologist and responsible clinical physicist.

Have standard cardiopulmonary resuscitation equipment directly available.

If any change in ICD functioning is observed, directly consult with the patient’s cardiologist to decide which steps should be taken next.

Monitor the ICD during the first months after radiotherapy. If functional changes are observed, consider replacement of the ICD.

We are aware of the rigorousness of the proposed guidelines. However, we believe that the inability to predict or detect the moment of ICD breakdown necessitates such actions. Note that only 11 ICDs were used in this study. Although this is the largest study to date that we know of, future studies incorporating more ICDs would be very useful to provide a more solid basis for guidelines for the irradiation of patients with an ICD.

In summary, the effect of radiation therapy on the newest generation of ICDs varies widely. If not programmed to monitoring only, the ICD might inappropriately give a shock during irradiation. Although most ICDs did not fail below 80 Gy, some devices had already failed at a very low dose of 0.5–1.5 Gy. Guidelines are formulated for the treatment of patients with an ICD.