RADIOISOTOPE POWERED CARDIAC PACEMAKERSFred N. Huffman, Ph.D., Joseph J. Migliore, M.D., William J. Robinson, B.S.

and John C. Norman, M.D.

Over 100,000 implantable cardiac pacemakers are currently being usedto rehabilitate patients with heart block. The chemical batteries power-ing these pacemakers usually fail within a few years. A variety of nuclearbatteries with the potential of providing long-lived (10-20 years) pace-maker power, are under development. This paper reviews the status of thisdevelopment. Nuclear powered pacemakers have reached the stage ofclinical evaluation. Their cost, although initially high, is not prohibitive.In our institution, third party insurance carriers have assumed their cost.The primary concern of widespread application is the maintenance of fuelcapsule integrity under all credible conditions.

PACEMAKER ENERGY SOURCESA wide variety of chemical, biologic and nuclear energy sources have

been explored to power cardiac pacemakers. Almost all currently im-planted pacemakers utilize primary batteries. Formerly, mercury-zinccells and electrodes produced an average pacemaker longevity of undertwo years. Contemporary electrodes require significantly lower energyinput to produce consistent pacing. As a result of reduced battery drainand improved circuits, commercially available pacemakers using primarybatteries are expected to average three to five years of longevity.' Sincethe average patient is 72 years at implantation, primary batteries arelikely to be adequate for the majority of clinical situations. Solid statelithium-halide batteries2'3 offer potential longevities significantly greaterthan zinc-mercury cells. Although the first clinical application of an im-plantable cardiac pacemaker4 used a nickel-cadmium battery, there hasbeen only limited clinical use of rechargeable systems due to operationalcomplexity, patient anxiety and lack of cells that would operate at bodytemperature. Recently, Fishell et al.5 addressed the latter problem anddeveloped a rechargeable pacemaker with an estimated lifespan of 20 years(Pacesetter Systems Inc.). Other investigators have sought to developbiogalvanic systems7'8 using body fluids as well as biological fuel cells.9While biological fuel cells should provide adequate power for pacing, nosuch unit has reached prototype evaluation.In addition to anxiety and expense, pacemaker replacement constitutes

a small but definite risk of infection and prolonged hospitalization. As aresult, batteries using the energy of radioisotope decay to provide long-lived (10-20 year) pacemaker power are being developed internationally.A genealogy of the more highly developed nuclear-to-electric conversiontechniques is included in Fig. 1. These conversion techniques can begrouped into two types: thermal and non-thermal. The thermal converters(whose output power is a function of a temperature differential) includethermoelectric and thermionic generators. The non-thermal converters

From the Cardiovascular Surgical Research Laboratories, Texas Heart Institute, of St. Luke'sEpiscopal and Texas Children's Hospitals, Texas Medical Center, Houston, Texas; The ThermoElectron Research and Development Center, Waltham, Massachusetts, and the CardiovascularDivision, Sears Surgical Research Laboratories, Harvard Medical School, Boston, Massachusetts.

52

(whose output power is not a function of a temperature difference) ex-tract a fraction of the incident energy as it is being degraded into heatrather than using thermal energy to run electrons in a cycle.

< ATOMIC ENERGY RESEARCH ESTABLISHMENT

Figure 1. Genealogy of Cardiac Pacemaker Power Sources

53

THERMAL CONVERTERS

Both thermionic and thermoelectric converters have been evaluatedfor pacemaker application. A thermionic converter,10 (Fig. 2), consists of ahot electrode which thermionically emits electrons over a space chargebarrier to a cooler electrode, producing a useful power output. Cesiumvapor is used to optimize the electrode work functions and provide anion supply (by surface contact ionization) to neutralize the electron spacecharge. A thermoelectric converter" connects thermocouples in series.Each thermocouple (Fig. 3) is formed by the junction of two dissimilarmaterials, one of which is heated and the other cooled. Metal thermo-couples have low thermal-to-electrical efficiency. However, the carrierdensity and charge can be adjusted in semiconductor materials such asbismuth telluride and silicon germanium to achieve much higher conver-sion efficiencies. A block diagram for thermoelectric and thermionic sys-tems is shown in Fig. 4.

Effective thermal insulation is essential to minimize heat loss withattendant reductions in fuel inventory, radiation exposure and expense.Both fiber and vacuum foil insulations are being used. Vacuum foil insula-tion, consisting of multiple layers of thin foils separated in a vacuum byoxide particles, provides lower thermal losses in a more compact unitthan does fiber insulation.

The dc-to-dc power conditioner is optional depending on the output po-tential of the converter. In principle, the voltage required (typically, aboutfive volts) can be obtained by connecting in series adequate numbers ofthermocouples or thermionic diodes. In practice, fabrication is often sim-plified by designing the converter for a low voltage output (approximately0.5 volts) and subsequently employing a dc-to-dc circuit to transform theconverter output voltage to the level required by the pacemaker elec-tronics.

The subsystem combinations are evident from Fig. 1. The Isomite ther-mionic converter'2 developed by McDonnell Douglas has been discon-tinued for pacemaker application. ARCO Nuclear'3 uses 528 metal (Cupron-Tophel) thermocouples in conjunction with vacuum foil insulation to ob-tain a high output voltage that can power a pacemaker without a dc-to-dcconverter. However, the low efficiency of the metal thermocouple resultsin a relatively high fuel loading. The development of this pacemaker hasbeen sponsored by the USAEC. CIT-Alcatel14 in France, the Atomic En-ergy Research Establishment15 in Great Britain, and the Gulf Energy andEnvironmental Systems Company in the United States all use low voltagebismuth telluride modules in combination with fiber insulations and dc-to-dc power conditioner units which decrease the relatively high thermo-electric conversion efficiencies. Nuclear Battery Corporation'6 uses a sim-ilar system, with the exception that vacuum foil insulation is substitutedfor fiber insulation. Their Atomcell has the lowest fuel loading (0.120 gram)of the thermal conversion systems. Siemans AG17 in West Germany vapordeposit 1420 bismuth telluride couples on a polyimide substrate to obtaina high potential thermopile. The thermocouples are connected electricallyin two series strings. Syncal Corporation18 is working on a high potentialsilicone germanium module in which all thermocouples are connected inseries-parallel.

54

COOLER COLLECTOR

CESIUM EMITTER9VAPOR

_-_ F,L////-//

HOTJUNCTION

VL

kCOLLECTOR

I/ AI/ EL

VL

Figure 2. Thermionic Converter with Potential Diagram.

COLD HOTJUNCTION JUNCTION

rfl_-

HEATOUT

Figure 3. Thermocouoe with Potential Diagran.

TISSUE

COLDJUNCTION

EL.

IpuEL.

_ HEAT PATH

--ELECTRICAL PATH

THERMAL RADIOISOTOPE POWERED PACEMAKER

Figure 4. Block Diagram of Radioisotope-Fueled Thermoelectric and Thermionic Systems.

55

HOTEMITTI

HEATIN

NON-THERMAL CONVERTERSNon-thermal converters extract a fraction of the nuclear energy as it is

being degraded into heat. Their outputs are not functions of temperaturedifferences as are thermoelectric and thermionic converters.Non-thermal generators can be grouped into three classes. The primary

generators consists of a condenser19 which is charged by the current ofcharged particles from a radioactive layer deposited on one of the elec-trodes. Spacing can be either vacuum or dielectric. Negatively chargedbeta particles or positively charged alphas, positrons or fission fragmentsmay be utilized. Although this form of nuclear-electric generator datesback to 1913, few applications have been found for the extremely lowcurrents and inconveniently high voltages provided by direct charginggenerators (Fig. 5).The betavoltaic cell is the most successful of several generators that rely

on secondary electronic excitation. One of the oldest is the ionic or con-tact potential battery which dates back to 1924 (Fig. 6). Two electrodesof dissimilar work function enclose a gas. Any source of ionizing radiationcan generate ion pairs in the gas, which are collected by the electric fieldset up by the contact potential owing to the dissimilar work functionelectrodes. The currents of oppositely charged particles flowing in oppositedirections add and may be passed through an external load to deliver anelectrical output. The U. S. Army sponsored work on this concept in theearly fifties at Ohmart Corporation and Tracerlab. Its drawbacks werepoor efficiency, instability and low power density.The irradiated P-N junction20 is a solid state analog of the ionic battery

(Fig. 7). Incident radiation (e.g. light, X-rays, alpha or beta particles)ionize hole-electron pairs. Minority carriers near the junction are collectedand constitute a current through an external load. Compared to the ionicbattery, efficiency is high, conversion volume is small and output is stable.Beta radiation of the P-N junction is the most practical. RCA was respon-sible for most of the early work on this betavoltaic generator. More re-cently, McDonnell Douglas2' has significantly improved the efficiency andpower output. Their betavoltaic unit is termed the "Betacel".

COLLECTORCHARGED PARTICLE F.L.DECAY ISOTOPE /

EMITTER COLLECTOR

VL

BETAEMITTER

VACUUM F L.ORDIELECTRIC

+ V -L

Figure 5. Direct Charging Generator with Potential Diagram

56

IONIZINGRADIATION 3

HIGH WORK LOW WORK tFUNCTION FUNCTION A_o

ELECTRODE o_@ ELECTRODE |

E)---w HIGH#

ARGON 4

VL

Figure 6. Work Function Battery with Potential Diagram.

..p.. I I " N"

IONIZING, RADIATION

F.L.7

VALENCEBAND

I I

I CONDUCTION

I 1777 7 7EL

Figure 7. Irradiated P-N Junction Battery with Potential Diagram.

IONIZINGRADIATION |

PHOTOVOLTAICCELL

Figure 8. Radioisotope-PhosphorliPhotovoltaic-Generator

57

F.L.

opts "N".

vm

VL

I

YLF- -

p

CZ

The phosphor-photovoltaic generator22 is an example of a dual conver-sion nuclear battery (Fig. 8). First, the radioactive energy is convertedinto light by a scintillating phosphor. Second, the photovoltaic cell con-verts the light into an electrical output. Theoretically, the phosphor-photo-voltaic generator has the potential of higher power density than the beta-voltaic generator. In practice, phosphor degradation has limited its appli-cation.

RADIOISOTOPE FUELSPlutonium-238 (Pu-238) is the unanimous choice of thermoelectric pace-

maker developers while most betavoltaic converters have utilized Prome-thium-147 (Pm-147). Plutonium-238 has a long half-life (89 years), reason-able power density (3.5 watts/cm3), low radiation output, establishedcontainment technology and low cost. Since Pu-238 is primarily an alphaemitter, its decay builds up a helium pressure inside the sealed fuel capsule.Most thermoelectric power sources (see Table 1) use a plutonium oxidefuel form. This refractory material (melting point of 4050°F) is chemicallyand biologically inert. To minimize the probability of particulate inhala-tion if the encapsulation is compromised, the fuel pellet is pressed and sin-tered. The dose rate can be further reduced by using high purity "medical-grade"' Pu-238 oxide23 enriched with Oxygen-16.The high energy and mass of the Pu-238 alpha particles would rapidly

degrade a P-N junction. Consequently, a low energy and mass beta particleis used for radiating a P-N junction to obtain an electrical output. Mostbetavoltaic investigations have used the inexpensive Promethium-147(Pm-147) beta emitter which has a half-life of only 2.62 years. Unfortun-ately, there does not appear to be a radioisotope with the requisite com-bination of beta energy, half-life and low dose rate that can be producedeconomically to take full advantage of this conversion technique. Althoughtritium24 and Nickel-63 have been considered, the beta range of the formeris quite short (resulting in excessive source absorption), while the latteris too expensive. Both give a drastically reduced electrical power densityrelative to the Pm-147 fueled betavoltaic converters. McDonnell-Douglasprojects a useful lifetime of 7 to 10 years for their Pm-147 fueled Betacelunits. Betavoltaic converters do not appear to have the potential for ex-tended life (up to 20 years) of the Pu-238 thermoelectric generators. Inthe case of either the Pu-238 or Pm-147 units, the integrity of the fuel con-tainment under all credible conditions is the primary design considera-tion.

NUCLEAR BATTERY COMPARISONOf the many nuclear conversion methods investigated for powering

pacemakers, only thermoelectric and betavoltaic systems have been ap-plied clinically and are commercially available. The subsystem combinationsof these power sources have been identified in Fig. 1. With a single excep-tion, all thermoelectric generators use semiconductor thermocouples toachieve a higher conversion efficiency than metal thermocouples. A sum-mary of the fuel, pacemaker electronics, cost and evaluation status is givenin Table 1.The tabulated nuclear power sources are comparable in size and weight

58

to chemical batteries. The fuel loadings of the semiconductor thermoelectricgenerators range from 0.120 to 0.200 gram of Pu-238; The ARCO Nuclearunit uses lower efficiency metallic thermocouples and requires a higherfuel inventory. The Nuclear Battery Corporation Atomcell is the mostefficient of the batteries, since it uses vacuum foil insulation as well asbismuth telluride thermocouples.The radiation dose rates from all the nuclear batteries are less than

20 mrem/hr at contact and less than 0.5 mrem/hr at 5 cm. Based on con-tinuing studies in our laboratories, the somatic effects to the patient fromthese dose rates should not be a deterrent to clinical use. Effects oncanines25 implanted with 16- and 24-watt Pu-238 sources (two orders ofmagnitude larger than those contained in nuclear-powered cardiac pace-makers) have been benign except for minor pathological alterations intissues within 6 mm of these sources. Radiation equivalent sources simu-lating the dose rate fields of 50 gram Pu-238 sources, but with negligiblethermal outputs, have been implanted in dogs26 for periods up to fouryears with similar results.The whole body, time averaged dose rate from a typical nuclear pace-

maker is approximately 0.5 rem/year compared to the whole body occupa-tional exposure limit of 5 rem/year recommended by the National Com-mittee on Radiation Protection.To date most nuclear pacemakers have been implanted in younger pa-

tients. At our institution, two ARCO Nuclear and four Medtronic(Laurens-Alcatel battery) nuclear pacemakers are being clinically evalu-ated. There have been no unusual results from these studies.

In summary, nuclear-powered cardiac pacemakers have reached thestage of clinical evaluation. Such pacemakers have the potential of longlife. Their cost, although initially high, is not prohibitive. Animal studiesindicate that the somatic and genetic radiation effects from a sealed sourceshould not preclude clinical use.The primary concern of widespread appli-cation is the maintenance of the fuel capsule integrity under all credibleconditions.

TABLE I

SUMMARY OF PRINCIPAL NUCLEAR PACEMAKERS

FuelRadioisotope Loading Pacemaker Ev,aluation

Facility Nation Fuel (GM) Electronics Cost Status C.mn,, nts

ARCO Nuclear U. S. Pu-238 Oxide 0.410 ARCO $ 5000/ 42 Clinical Cases Developnunt sponsored by(0-16 Enriched) Pacemaker USAEC

"Atomcell" U. S. Pu-238 Oxide 0.120 - $ 2500/ Animal Evaluation ost efficient of available(,Nuclmear Battery Corp.) (0- 16 Enriched) Battery n-clear bvatteri,sAtomic Energy Great Pu-238 Nitride 0. 180 - _ -80 Clinical Cases Based on RIPPLE techn7,l7gyResearch Estab. Britain(AERE)

"Btacel 400" U. S. Pm- 147 Oxide 0. 105 Biotronik ( 2375/ 60 Clinical Ca.sc 1'otential life I1

ha that(McDonnell Douglas IF;P- S1; I 'aP emas.k"r)f P'l- Z 38 unit,Gulf Energy and U. S. Pu-238 Oxide 0. 180 - Ss Z:0oo/ Animal Fsaluati.n Basi-Ily if 1ential to AFbEnvironmental Systemls (0- 16Enriched) B3attery >a tt,- ry

CIT-Alcatel France Pu-238/ 0. 160 Medtronics $ 5000/ 400 Clinical C- First pa-a17k,r to e usedSc Alloy Model 9000 Psac-mak- linically

Siemens AC. West Pu-238 Oxide 0.200 - - - e1ap-)at. d tEin-filn,Germany th rmopile

Sync,al U. S. Pu-238 Oxide - - rototype th-n-pil- Direct highi -ol.tage -ithunder t, st siIico:n-gern,.niu., th--r,).pil,-

59

REFERENCES

1. Furman S: Cardiac Pacing-An Endless Frontier. Medical Instrumentation 7:169, 19732. Greatbatch W, Lee JH, Mathais W, et al: The Solid State Lithium Battery: A new improved

chemical power source for implantable cardiac pacemakers. IEEE Trans. Biomed. Eng.BME-185: 317, 1971

3. Fester K, Doty RL: Solid-state batteries for cardiac pacemakers. Medical Instrumentation7:172, 1973

4. Greatbatch W, Chardack W: A transistorized implantable pacemaker for the long-termcorrection of complete atrio-ventricular block. Proc New England Research and Eng Meeting(NEREM) 1:8, 1959

5. New pacemaker is rechargeable. Electronics: 35, 19736. Barnhart PW, Fischell RE, Lewis KB, et al: A fixed-rate rechargeable cardiac pacemaker.

Applied Phys Lab Tech Dig 9: 2, 19707. Cassel J, Satinsky V, Eibling D, et al: Implanted silver-silver chloride magnesium power

sources. Medical Instrumentation 7: 176, 19738. Schaldach M, Kirsch V: In vivo electrochemical power generator. Trans Amer Soc Artif Int

Organs 16: 184, 19709. Wolfson SD, Yao SJ, Geisel A, et al: A single electrolyte fuel cell utilizing permselective

membranes. Trans Amer Soc Artif Int Organs 26: 193, 197010. Direct generation of electricity. Edited by KH Spring. London-New York, Academic Press,

1965, page 19611. Heikes RR, Ure RW Jr: Thermoelectricity: science and engineering. New York-London,

Interscience Publishers, 196112. DeSteese JG: Development of thermionic radiostopic batteries. Proceedings of Second Inter-

national Symposium on Power from Radiosotopes, Madrid: 339, 197213. Maurer GW, Hursen TF, Kolenik SA: Radioisotope powered cardiac pacemaker development

program. Trans. American Nuclear Society 14: 505, 197114. Alais M, Etieve B, Stahl A, et al: The Gipsie Radiosotopic Generator for use in cardiac

pacemakers. Proceedings of Second International Symposium on Power from Radiosotopes,Madrid: 191, 1972

15. Myatt J, Brown MH, Neighbor F, et al: Radiosotope fueled batteries using Plutonium-238and Strontium-90. Proceedings of Second International Symposium on Power from Radiosotopes,Madrid: 397, 1972

16. Greatbatch W, Bustard TS: A Pu-238 dioxide nuclear power source for implantable cardiacpacemakers. IEEE Transactions on Biomedical Engineering BME-20: 332, 1973

17. Renner T, Rittmayer G, Falkenberg D, et al: An isotopic thermoelectric battery for cardiacpacemakers. Proceedings of Second International Symposium on Power from Radiosotopes,Madrid: 207, 1972

18. Private Communication, Mr. Valvo Raag19. Voinon M, Tannenberger H: Possible performances of a tritium cell. Proc. of Second Inter-

national Symposium on Power from Radiosotopes, Madrid: 311, 197220. Argrist SW, Direct energy conversion. Boston, Allyn and Bacon, 1965, page 16721. Olsen LC, Seeman SE: Betacel batteries for biomedical applications. Proceedings of Eighth

Intersociety Energy Conversion Engineering Conference, Philadelphia: 454, 197322. Schalch D, Scharmann A: Practical limits of radiophotovoltaic conversion systems. Proceed-

ings of Second International Symposium on Power from Radioisotopes, Madrid: 253, 197223. Mullins LJ, Matlock GM, Bubernak J, et al: Characterization and properties of medical

grade Pu-238 fuels. Proceedings of second International Symposium on Power from Radio-isotopes, Madrid: 49, 1972

24. Gott FC: A tritium nuclear cardiac pacemaker battery. Proc. of Second International Sym-posium on Power from Radiosotopes, Madrid: 325, 1972

25. Huffman FN, Molokhia FA, Norman JC: Thermal and radiation effects of Pu-238 fuelcapsules in dogs and primates. Trans. American Nuclear Society 14: 510, 1971

26. Sandberg GW, Huffman FN, Norman JC: Experimental observations of intracorporeal Stron-tium-90-Americium 241 Beryllium sources simulating radiation fields from nuclear poweredartificial hearts. Annals of Thoracic Surgery 9: 401-409, 1970

27. Norman JC, Sandberg GW, Huffman FN: Current concepts. Implantable nuclear-poweredcardiac pacemakers. N Engl J Med 283:1203-1206, 1970

60