Internal Radionuclide Radiation Dosimetry: A Review of Basic Concepts and Recent Developments *

Pat B. Zanzonico

Nuclear Medicine Service, Memorial Sloan-Kettering Cancer Center, New York, New York

Relatively low administered activities yield important diag Internal dosimetry deals with the determination of the amount nostic information, the benefits of which far outweigh any and the spatial and temporal distribution of radiation energy potential risk associated with the attendant normal-tissue deposited in tissue by radionuclides within the body. Nuclear radiation doses. Such small risk-to-benefit ratios have been medicine has been largely a diagnostic specialty, and model very forgiving of possible inaccuracies in dose estimates. derived average organ dose estimatesfor risk assessment,the By incorporation of radionuclides in appropriately large traditionalapplicationof the MIRDschema,have provenentirely adequate. However,to the extent that specific patients deviate amounts into target tissue—avidradiopha.rmaceuticals, a kinetically and anatomically from the model used, such dose sufficiently high radiation dose may be delivered to produce estimates will be inaccurate. With the increasing therapeutic a therapeutic response in tumor or other target tissue. With application of internal radionuclides and the need for greater escalating administered activities and associated normal accuracy, radiation dosimetry in nuclear medicine is evolving tissue doses, serious radiation injury can ensue, however. It from population- and organ-average to patient- and position becomes imperative, then, that the magnitude of the target specific dose estimation. Beginningwith the relevant quantities tissue and at-risk normal tissue radiation doses be estab and units, this article reviews the historicalmethods and newly developedconceptsandtechniquesto characterizeradionuclide lished with reasonable accuracy and precision and used in radiationdoses. The latter includethe 3 principalapproachesto conjunction with reliable dose—responserelationships for the calculation of macroscopic nonuniform dose distributions: target tissues and dose—toxicity relationships for normal dose point-kernelconvolution,MonteCarlosimulation,andvoxel tissues. With the ongoing development of new radiopharma S factors. Radiationdosimetryin “sensitive―populations,includ ceuticals and the increasing therapeutic application of inter ing pregnantwomen, nursingmothers,and children,also will be nal radionuclides (particularly in the form of radioimmuno reviewed. therapy), radiation dosimetry in nuclear medicine continues KeyWords:dosimetry;MIRD;radionuclidetherapy to evolve from population- and organ-average to patient J NucIMed2000;41:297—308 and position-specific dose estimation (1—3).Patient-specific dosimetry refers to the estimation of radiation dose to tissues of a specific patient, based on his or her individual body habitus and measured radiopharmaceutical kinetics rather ntemal radionuclide radiation dosimetry deals with the than on an average anthropomorphic model and hypothetic determination of the amount and spatial and temporal kinetics. In contrast to the average dose, position-specific distribution of radiation energy deposited in tissue by dosimetry refers to radiation doses to specific points in a radionuclides within the body. Internal dosimetry has been tumor or organ and thus reflects the spatial variation in dose applied to the determination of tissue doses and related within a target tissue. This article reviews the historical quantities for occupational exposures in radiation protection, methods and newly developed concepts and techniques to environmental exposures in radiation epidemiology, and characterize radionuclide radiation doses. diagnostic and therapeutic exposures in nuclear medicine. Historically, nuclear medicine has been largely a diagnostic STOCHASMAND DETERMINISM specialty, and the associated risk—benefitanalyses implicitly performed by the clinician have been straightforward. Dosimetric quantities may be characterized as stochastic or deterministic (i.e., nonstochastic). A quantity subject to statistical fluctuations because of the small number of Received Sep. 20, 1999; revision accepted Oct. 12, 1999. For correspondenceor reprintscontact:Pat B. Zanzonico,PhD, Nuclear energy deposition events contributing to the quantity or the MedicineService,MemorialSloan-KetteringcancerCenter,1275YorkAve., small volume of the region for which the quantity is being NewYork, NY 10021. *NOTE: FOR CE CREDIT, YOU CAN ACCESS THIS ARTICLE ON THE determined is termed “stochastic.―The mean, or expecta SNMWEBSITE(httpJ/www.snm.org)UNTILAUGUST2000. tion value, of a large number of determinations of a

RADIATION Dosmtirn@y •Zanzonico 297 stochastic quantity is termed “deterministic.―Each stochas unit is the rad (1 rad 100 erg/g); 1 Gy equals 100 rad, and tic quantity thus has a corresponding, generally more 1 rad equals 1 cGy (or 10 mGy) (11). familiar, deterministic quantity. The field of microdosimetry deals with the number, size, LinearEnergyTransferandLinealEnergy and spatial and temporal distributions of individual energy The quality and the quantity of radiation are important deposition events, particularly in submicroscopic structures determinants of the frequency or severity of radiogenic (4—6).Increasingly, radiation dosimetry in nuclear medicine biologic effects. The quality of a radiation is related to the deals with the nonuniformity of dose within organs, among characteristics of the microscopic spatial distribution of cells, and even within cells. Even so, dose is still almost energy-deposition events. Sparsely ionizing radiations, such always expressed in terms of deterministic quantities, and as x- and ‘y-raysand intermediate- to high-energy electrons such dosimetry should therefore not be referred to as and @3-rays,are characterized as low-quality radiations. microdosimetry. Terms such as “small-scaledosimetry― or Densely ionizing radiations, such as low-energy electrons “cell-leveldosimetry― would be more appropriate. How (e.g., Auger electrons), protons, neutrons, and a-rays, are ever, the nonuniformity of energy deposition associated with typically characterized as high-quality radiations. For the a-ray emitters (7,8) and with radioimmunotherapy (9) may same absorbed dose, the frequency or severity of biologic warrant the rigorous application of microdosimetry (10). effects is generally less for sparsely ionizing than for densely ionizing radiations. RADIATIONQUANTiTiESAND UNITS The quality of radiation is characterized by the linear energy transfer (L or LET): Definitions of various quantities used to specify radiation dose and of selected related quantities are presented here. A dE compilation of System Internationale (SI) and conventional Lank, Eq.3 quantities and their symbols, units, and conversion factors are presented elsewhere (6,11). where dE = the energy lost by a charged particle (or the secondary charged particle produced by the primary radia Administered Activfty tion) in traversing a distance in matter and dl = the distance Administered activity is specified by the SI unit bec traversed in matter. querel, equaling 1 disintegration/s (dps) or some multiple LET is the deterministic correlate of the stochastic thereof. The conventional unit of activity, the Curie, corre quantity, lineal energy (y): sponds to 3 X l0@@dps. Note that 37 MBq equals 1 mCi and €1 37 kBq equals 1 MCi. It is important to distinguish the yan--, Eq. 4 radiation dose from the administered activity, with the former quantity taking into account the radionuclide and its where €1= the energy imparted to matter by a single physical characteristics, tissue mass, and the radiopharmaceu energy-deposition event and 1 = the mean chord length of tical and its kinetics as well as the administered activity. the volume of matter. The unit of linear energy transfer and lineal energy is the Absorbed Dose and Specific Energy same: the SI unit is the J/m and the conventional unit is the Perhaps the most widely used and biologically meaning keV4tm; 1 J/m equals 6.25 X 10@keV/j.im, and 1 keV/@.im ful quantity for expressing radiation dose, the absorbed dose equals 1.60 x l0@°J/m(11). (D), is defined as:

dE RelativeBiologicEffectiveness Dan—, Eq. 1 The influence of LET on the frequency or severity of dm biologic effects is quantified by the relative biologic effective where d@= the mean energy imparted by ionizing radiation ness(RBE): to matter and dm = the mass of matter to which the energy is RBE(A) an Drricrence imparted. Eq.5 The absorbed dose, defined for all ionizing radiations in any stopping medium, is the deterministic correlate of the where Dref@n@= the absorbed dose of reference radiation stochastic quantity, specific energy (z1): (typically a widely available, sparsely ionizing radiation, such as cobalt-60 “1'-rays)required to produce a specific, €1 z1 Eq.2 quantitatively expressed biologic effect and DA = the m absorbed dose of radiation A required to produce the same where €@= the energy imparted to matter by a single frequency or severity of the same specific biologic effect, energy-deposition event and m = the mass of the matter. with all pertinent parameters maintained as nearly identical The unit of absorbed dose and specific energy is the same: as possible. The RBE is a ratio of absorbed doses and thus is the SI unit is the gray (1 Gy = 1 J/kg), and the conventional a dimensionless quantity.

298 [email protected]•Vol.41 •No. 2 •February2000 RadiationWeightingFactorandEquivalentDose total-body irradiation (Table 2). The effective dose is similar A simplified version of the RBE, the radiation weighting in concept to the effective dose equivalent (HE) introduced factor (wR), was devised for purposes of radiation protection previously by the International Commission on Radiation (Table 1). The so-called equivalent dose (HT) in tissue or Units and Measurements (12) and the National Council on organ T is related to the radiation weighting factors (w@Jand Radiation Protection (13) and representing a single-value the mean absorbed doses (D,@.,R)to tissue or organ T from estimate of the net harm from any low-dose (e.g., diagnostic radiations (R): nuclear medicine) exposure. However, the dose equivalent is Eq 6 based on the absorbed dose at a point in tissue weighted by @ HT wRDTJ@. the LET-dependent distribution of quality factors at that point. The equivalent dose, in contrast, is based on the The wR and H1 are similar, respectively, to the older average absorbed doses in the tissue or organ weighted by quantities of quality factor (Q) and dose equivalent (H). The the radiation weighting factor for the radiation actually key difference is that the former are related to the mean dose impinging on that tissue or organ. The effective dose and the to a tissue or organ, whereas the latter are related to the dose effective dose equivalent do not apply to, and should not be at a point and thus are often less useful. used for, high-dose (e.g., therapeutic nuclear medicine) Tissue Weighting Factor and Effective Dose exposures (14). The effective dose (E) is intended to provide a single The unit of equivalent dose and effective dose is the same: value estimate of the overall stochastic risk (i.e., the total the SI unit is the sievert and the conventional unit is the rem; risk of cancer and genetic defects) of a given irradiation, 1 Sv equals 100 rem, and 1 rem equals 1 cSv (or 10 mSv) whether received by the whole body, part of the body, or 1 or (13). wR and w1 are dimensionless quantities (Tables 1 and more individual organs: 2) (15,16). Ean@wTHT Eq.7 THE TRADITiONALMIRDSCHEMA @ = wTwRDT,R, Eq. 8 The methodology now widely used used for internal dose calculations in medicine, including age- and sex-specific where wT the weighting factor for tissue or organ T, a reference data for human anatomy and body composition, dimensionless quantity representing the fraction contributed was developed by the MIRD committee of the Society of by tissue or organ T to the total stochastic risk (i.e., the Nuclear Medicine and is generally referred to as the MIRD combined total risks of cancer or of severe hereditary defects schema or MIRD formalism (2,17—19).The International for all subsequent generations) as the result of a uniform, Commission on Radiological Protection (ICRP) has devel o_ a similar methodology and similar reference data (20). TABLEI The MIRD schema, including notation, terminology, WR for Calculation of HT to Tissue or Organ mathematic methodology, and reference data, has been disseminated in the form of the collected MIRD pamphlets Typeandenergyrange and associated publications (2, 17—19).With the publication xandy-rays,electrons,positrons,andmuons* 1 of work by Christy and Eckerman (21), age- and sex-specific Neutrons,energy body habiti other than the original 70-kg adult anthropomor <10 keV 5 phic model (known as “StandardMan―)(22) now are 10-100 keV 10 >100 keV-2 MeV 20 incorporated into the MIRD schema. In addition, several >2-20 MeV 10 computerized versions of the MIRD schema have been >20 MeV 5 developed, including MIRDOSE (23), DOSCAL (which Protonstotherthanrecoilprotonsandenergy> 2 MeV 20 incorporates mean tumor doses) (24), and MABDOS (with a particles, fission fragments, nonrelativistic heavy nuclei curve fitting and modeling features) (25—27). In its traditional application to diagnostic radiopharmaceu *ExcludingAuger electronsemittedfrom nucleiboundto DNA, ticals, the MIRD schema implicitly assumes that activity and becauseaveragingdose in this case is unrealistic.Techniquesof microdosimetryare moreappropriateinthiscase. cumulated activity are uniformly distributed within organ tIn circumstancesin whichhumanbody is irradiateddirectlyby size source regions and that radiation energy is uniformly >100-MeVprotons,RBEis likelyto be similarto that of low—lineardeposited within organ size target regions. Moreover, dosim energytransferradiation,and,therefore,wRof approximatelyunity etry for diagnostic radiopharmaceuticals is generally based wouldbe appropriateforthatcase. 1@VRvalue for high-energy protons recommended here is lower on (a) average time—activitydata in animal models or in than that recommendedby InternationalCommissionon Radiologi small cohorts of human subjects and (b) age- and sex calProtection(1991). specific “average―models of human anatomy. The tradi All values relate to radiationincidenton body or, for internal tional MIRD schema does not incorporate tumors as either sources,emittedfromsource.Adaptedwithpermissionof(16). source or target regions.

RADIATION DOSIMETRY •Zanzonico 299 TABLE2 source and target regions, it can be used to calculate the dose WT for Tissue or Organ for Calculation of Effective Dose to virtually any structure from virtually any activity distribu tion, from microscopic to macroscopic to whole organ and wT Tissue or organ whole body. With the publication of MIRD Cellular S 0.01 Bonesurface,skin Factors (28), the MIRD schema has been extended to 0.05 Bladder,breast,liver,esophagus,thyroid,remainder* cellular and subcellular source and target regions, with the 0.12 Bonemarrow,colon,lung,stomach 0.20 Gonads cell and the nucleus modeled as unit density concentric spheres of radius 3—10and 1—8pm, respectively. More recently, MIRD pamphlet no. 17, The Dosimetry of Nonuni *For purposes of calculation,remainderis composedof the form Activity Distributions: Radionuclide S Values at the followingadditionaltissuesand organs:adrenals,brain,smallintes tine, largeintestine,kidney,muscle,pancreas,spleen,thymus,and VoxelLevel(29) has extended the MIRD schema to arbitrary uterus.Listincludesorgansthatare likelytobe selectivelyirradiated. macroscopic activity distributions in 3 dimensions for Someorgansin listare knownto be susceptibleto cancerinduction. calculation of the resulting macroscopic dose distribution. In Ifothertissuesandorganssubsequentlybecomeidentifiedas having this context, macroscopic refers to volume elements (or significantriskof inducedcancer,they will then be includedeither voxels) 3 mm or greater in dimension. In both publications, with a specificw, or in this additionallist constitutingremainder. Remaindermay also include other tissues or organs selectively the distance-dependent photon and electron dose contribu irradiated.In thoseexceptionalcasesinwhich1 remaindertissueor tions are included and, in the case of @3-rays,the actual organreceivesequivalentdosein excessof highestdosein any of energy spectra (versus simply the mean energy) are used. the 12organsforwhichweightingfactorisspecified,weightingfactor of0.025 shouldbeappliedtothattissueororganandweightingfactor Patient-SpecificDosimetry of0.025 to averagedoseinotherremaindertissuesororgans. Individual variations in radiopharmaceutical kinetics may Valueshave been developedfor referencepopulationof equal result in substantial deviations from population-average numbersof both sexes and wide range of ages. In definitionof dose estimates, so that they are not predictive of target tissue effectivedose,they applyto workers,to whole population,and to eithersex.ThesewTvaluesarebasedon roundedvaluesof organ's response or normal tissue toxicity in radionuclide therapy. contributionto totaldetrimentAdaptedwithpermissionof(16). This is illustrated by the various dosimetric approaches to radioiodine treatment of hyperthyroidism (30). The clinical effectiveness of radioiodine treatment of hyperthyroidism is presumably related to the absorbed dose BEYONDTHE TRADITIONALMIRDSCHEMA to thyroid follicular cells. Yet, the most common dose The traditional MIRD schema has proven invaluable for prescription algorithm, a fixed administered activity, entails dosimetric risk assessment in diagnostic nuclear medicine. administration of a specified activity (currently, approxi However, to the extent that specific patients deviate kineti mately 370 MBq [10 mCi]) to all hyperthyroid patients. By cally and anatomically from the respective kinetic and ignoring those highly variable individual parameters (i.e., anatomic averages, tissue dose estimates will be inaccurate. thyroid uptake, biologic half-life, and mass) that determine Robert Loevinger, an originator of the MIRD schema, has absorbed dose, it is unlikely that there will be a correlation stated that, “.. .there is in principle no way of attaching a between the administered activity and the absorbed dose numerical uncertainty to the profound mismatch between (Fig. 1). Although up to 85% of patients will eventually be the patient and the model (the totality of all assumptions that “cured―(i.e., made euthyroid or hypothyroid) by the enter into the dose calculation). The extent to which the administration of single doses of standard amounts of model represents in some meaningful way a patient, or a radioiodine (and presumably 100% of patients treated with class of patients, is always open to question, and it is the repeated administrations) (31), such an arbitrary approach responsibility of the clinician to make that judgment―(19). does not permit individually optimized therapy and would Because of the large risk—benefitratios associated with not identify patients such as “small-pool―patients for whom diagnostic radiopharmaceuticals, any inaccuracies (perhaps radioiodine therapy may be inappropriate or even hazardous. up to 100% [2]) in tissue absorbed dose estimates for Prescription of a specified activity concentration (typi individual patients probably are unimportant. With therapeu cally 2000—3000 MBq/g [55—80pCi/g]) in the thyroid tic radiopharmaceuticals, however, the risk-benefit ratios are incorporates 2 important dosimetric variables, thyroid up dramatically smaller, and the tolerances for inaccuracies in take and mass, into calculation of the administered activity dose estimation are greatly reduced. but implicitly assumes a constant thyroid half-life of radio With the growth of radionuclide therapy, various tech iodine: niques beyond the traditional MIRD schema have been developed to improve the accuracy of dose estimates, administered activity = including techniques for patient- and position-specific radia prescribed activity concentration tion dosimetry. Any perception that such techniques some x glandmass(g) X 100 how correct inherent limitations of the MIRD schema is .Eq.9 24-h % uptake mistaken, however. Indeed, a remarkable strength of the MIRD schema is its generality: by judicious selection of A prescribed absorbed dose (Gy [rad]) is the dose

300 Tiii@Joui@u.. OF NUCLEARMEDICINE•Vol. 41 •No. 2 •February 2000 100

I a

FIGURE1. Relativeerrorin calculated ‘31labsorbeddoseto thyroidasfunctionof relative error in thyroid 24-h uptake, bio logic half-life, and mass. Baseline param eters are 24-h uptake of 25%, biologic half-lifeof9Od, and massof2l g. Notethat, for errors in biologic half-life, relative error -60 inthyroidabsorbeddosebecomesimpor tant only for very large negativeerrors. For -80 example,if actual biologichalf-lifewere only 5 d (typical of small pool patients), -100 actualthyroiddosewouldbe overestimated S Errer In Pamms@r by—60%if oneassumesstandardbiologic half-lifeof90 d.

prescription<10%requiresalgorithm presently used least often, because itulation.) Because the 1311@y-raystypically contribute assumedconsuming.serial uptake measurements and is therefore timeof the dose, deviations of thyroid masses from the doseincorporatingHowever, it is the most logical dosimetrically,value of 25 g have only a minor effect on the thyroid 9.eters all pertinent and practically evaluable param calculated using Equation for.into the calculation of the administered activity:The following have been offered as general guidelines

. . . radioiodine therapy of hyperthyroidism (35,36): administered activity (kBq) =single-dose . . Gravesprescribed for young patients and those with uncomplicated absorbed dose (cGy)disease (i.e., small glands and mild to moderate hyperthyroid x glandmass(g) X 6.67 x 37 7,000—8,000rad; for patients with complicated Graves' @ (Tl@)eff(day) X 24-h % uptake ‘ 10ism),disease (i.e., larger glands and more severe hyperthyroid nodularwhere ism), 10,000—12,000rad; and for patients with toxic typicalthyroid,(T1@)eff= the effective half-life of radioiodine in thegoiter, 15,000—25,000 rad. However, in contrast to which in turn equalshyperthyroid patients ([T1,@}b= 20—25d), 15% of hyperthy ofEq. roid patients have a more rapid thyroidal turnover (T1,@)@X (Tl,@)b 11radioiodine ([Tl,@]b= 5—10d). This is thought to be the result awhere (T1,@)@+ (Tl,@)b' of a small thyroidal iodine pool (small-pool syndrome). As ashalf-life(Tl,@)b the biologic (or radioactive decay-corrected)result, serum protein-bound radioiodine, the source of (tophysicalof radioiodine in the thyroid, and (T1,@)@= themuch as 90% of the 1311blood absorbed dose, is elevated serum),thehalf-life of ‘@‘I(8.04 d). Implicit in Equation 10 areas high as 2% of the administered activity per liter of thetheassumptions that the only significant absorbed dose towith a resulting increase in the blood absorbed dose. At becausetionsthyroid is from self-irradiation, all nonpenetrating radia same time, the thyroid absorbed dose is reduced hasabsorbed(i.e., @3-rays)emitted within the thyroid are completelyradioiodine is more rapidly secreted from, and therefore smallable locally, and the proportion ofthyroid mass attribut less time to irradiate, the thyroid. The relatively probablyderivedto lymphocytes is negligible. The factor of 6.67 wasthyroidal absorbed dose in small-pool patients is by Marmnelli et al. (32—34)(a) assuming the thyroidresponsible for at least some failures of radioiodine treat to7spheres,have a mass of 25 g and to consist of 2 unit density tangentment. Barandes et al. (37), for example, reported a series of 7000-yieldingeach of mass 12.5 g and radius (R) of 1.44 cm,small-pool hyperthyroid patients in whom a standard the(b) a mean geometric factor (gsp,@re)3 1TR of 12.3 cm;cGy (7000-md) thyroid absorbed dose would require tonentialassuming radioiodine in the thyroid follows a monoexpo administration of 1040 MBq (28 mCi) 1311(in contrast time—activity function; (c) using a mean @3-rayonly 120 MBq (3.2 mCi) typically required) and theoreti a0.00223energy of 0.191 MeV and a specific rny-rayconstant ofcally deliver a blood absorbed dose of 150 cGy (150 rad), benignincorporatingR cm2/mCi/h (in conventional units) for 1311;and (d)prohibitively high incidental dose for treatment of a (Thedisease.factorthe appropriate unit conversion factors. isto for converting pCi to kBq, 37, is presented explicitlyA general patient-specific treatment-planning paradigm retain the familiar factor of 6.67 in the Marineffi form as follows (30,38—42).A tracer (diagnostic) amount of the

RADIATIONDOSIMETRY•Zanzonico 301 therapeutic radiopharmaceutical is administered to the pa Potential radiogenic damage to the hematopoietic bone tient. Serial time—activity measurements are performed for marrow is the primary dose-limiting toxicity for systemic blood, tumor, or other target tissue; critical normal organs; or radionuclide therapy in general and radioimmunotherapy in the total body (30,41,43,44). These kinetic data are inte particular (45,97,98). A variety ofapproaches have therefore grated to determine the corresponding cumulated activities been pursued in an effort to establish a predictive dose (or residence times), and the absorbed doses per unit response relationship for myelotoxicity in radioimmuno administered activity are calculated. The actual therapeutic therapy (97,99—101). Although no such correlation has administered activity is that projected to deliver maximum proven especially useful (50,102—107),absorbed dose yields tolerated doses to 1 or more critical normal tissues or, less a better correlation than administered activity. Moreover, commonly, a minimum effective dose to tumor or other marrow absorbed dose appears to be a marginally better target tissue. For ‘31I-iodidetreatment of metastatic thyroid predictor of myelotoxicity than whole-body absorbed dose. cancer, for example, the therapeutic administered activity is However, in an intermediate absorbed-dose range, myelotox that calculated to deliver no more than 2 Gy (200 rad) to icity has been unpredictable. This may be a notable example blood (as a surrogate for bone marrow) (30,41,45,46). In of the inadequacy of the average dose as a quantitative radioimmunotherapy of non-Hodgkin's B-cell lymphoma descriptor of tissue irradiation and potential toxicity (Fig. 2). with 1311-labeledanti-B! (anti-CD2O) monoclonal antibody, Nonuniformity of dose in target tissues such as tumor on the other hand, the therapeutic administered activity is likewise may make it difficult to reliably predict therapeutic that delivering a dose of 0.75 Gy (75 rad) to the total body response in radionuclide therapy. O'Donoghue (108) has (again as a surrogate for bone marrow) (47—50). For modeled the impact ofdose nonuniformity (109) on radiocur radiolabeled monoclonal antibody and other targeted thera ability of tumors. As shown in Figure 3, tumor response is pies in which uptake is saturable or otherwise nonlinear, poorer (i.e., tumor cell survival is greater) as dose nonunifor with varying “mole―amounts administered, mathematic mity increases. The dose—responsecurve is concave upward (e.g., compartmental) modeling of the tracer kinetic data (Fig. 3D), indicating that the tumor-sparing effect of dose may be useful in determining the optimum amount (mole or nonuniformity is greatest at higher doses. A substantial mg) as well as activity of the therapeutic radiopharmaceuti fraction of tumor cells will therefore receive sublethal doses, cal (51—57). After the therapeutic administration, serial and the tumor may therefore not regress, even if the average time—activity measurements and cumulated-activity and tumor dose is sufficiently high to otherwise create an absorbed-dose calculations may be repeated and the pro expectation of a significant therapeutic response. A clinically jected and actual therapeutic absorbed doses compared. predictive average dose—responsecurve may be difficult or NonuniformDoseDistributions impossible to derive in the face of such nonuniformity. Normal tissue toxicity and tumor therapeutic response There are at least 3 approaches to the calculation of may not correlate with average doses, even when based on macroscopic nonuniform dose distributions (29): dose point individualized kinetics. Thus, in addition to patient-specific kernel convolution, Monte Carlo simulation, and voxel S dosimetry, the issue of spatial nonuniformity of dose has factors. The dose point-kernel is currenfly the most widely become increasingly important at the macroscopic (29,58— used of these approaches (42,60—63,73,110), primarily be 77) and microscopic (78—96)levels. cause of the demanding computational requirements of

UFem Head OHum Head @L3+L4 3.0

0' 2.5 I C, 2.0

1.5 FIGURE2. Nonuniformityof bonemar 0 rowabsorbeddose.Absorbeddoseperunit administered activity (mGy/MBq) to bone 1.0 marrowinfemoralheads(FernHead),hu meralheads(HumHead),andlumbarverte 0.5 brae 3 and 4 (L3+L4) was evaluatedin 9 2 patients with leukemia who received 300 111 MBq131l-labeledHuM195anti-CD33mono iii:71I TiIi:1 clonalantibody.Note2- to3-foldvariationin estimated absorbed doses among 3 mar 1 2 3 4 5 6 7 8 9 row sites. (Adapted with permissionof PatientNumber (131).)

302 THE JOURNALOF Nuci..@ MEDICINE•Vol. 4! •No. 2 •February 2000 A 0.06 B Heterogeneity

005 0.2 . 1.E+00 :@‘0.04 @ C I.E-02 10.03 @ 0.02 ‘@1.E-04

0.01 [email protected]

0 @B1.E-08 0 20 40 60 80 Dose (Gy) 1.E-10

@ C .2E-6 MeanDose (Gy) D 0 20 40 60 1.OE-6 1

@‘8.OE-7

@ 6.OE-7 i@ o.i

@ 4.OE-7 I@

2.OE-7 -@ 0.01 0.OE+0 Os 0 20 40 60 80 Dose (Gy) 0.001

FIGURE3. Effectof dosenonuniformityontumorresponse(108).(A) Hypotheticnonuniformdoseto tumorcellpopulation representedby normaldistribution(109) with averageof 40 Gy,SD of 7 Gy,and fractionalSD (FSD)of 7/40 Gy = 0.175. (B) Tumor cell survivalprobabilityfor dose distributionin (A), assumingmonoexponentialtumorcell survivalcurvewith meanlethaldose (D0)of 2.85 Gy (i.e., a = 0.35/Gy).Overalltumorcellsurvivalfractionis representedby area undercurve.(C) Overalltumorcellsurvival fraction as function of dose nonuniformity expressed as FSD of average dose from 0 (i.e., uniform dose) to 0.5 and of average tumor dose from 10 Gy (highest curve) to 60 Gy (lowest curve). Tumor cell survival is greater as dose nonuniformity increases. (D) Dose—responsefordosenonuniformity(i.e.,FSD)of0.35, correspondingto pointsintersectingdottedverticalline,is concaveupward. (Adaptedwith permissionof( 108).)

Monte Carlo simulation and the limited availability of voxel ing, however (58). Tabulations of voxel S factors, conceptu S factors. (A dose point-kernel is the radial distance ally equivalent to voxel source-kernels (the mean absorbed dependent absorbed dose about an isotropic point source in dose to a target voxel per radioactive decay in a source an infinite homogeneous medium [typically a soft tissue voxel, both of which are contained in an infinite homoge equivalent medium such as water].) With the wider availabil neous soft-tissue medium) (76, 77), are becoming available ity of high-speed desktop computers and of compatible (29). In contrast to techniques based on the dose point-kernel simulation codes, the use of Monte Carlo analysis may and Monte Carlo simulation, the voxel S factor method does increase (58,59, 111). Monte Carlo—based dosimetry can not require specialized computer facilities and is relatively more accurately account for tissue variations in mass density fast. Thus, it may emerge as the practical method of choice and atomic number as well as edge effects, which may be for calculation of macroscopic nonuniform dose distribu important at the periphery of the body and at soft tissue—lung tions. and —boneinterfaces (58). For example, if the relevant Once a dose distribution has been calculated, a correspond microscopic distribution data were somehow available (e.g., ing dose—volumehistogram can be derived. A dose—volume by autoradiography of biopsy specimens), Monte Carlo histogram is a graph of the fraction of the tumor or organ analysis might be especially applicable to normal lung volume receiving a specified dose versus the dose (differen dosimetry in radioiodine treatment of metastatic thyroid tial form) or the fraction of the tumor or organ volume cancer, particularly in dosimetrically problematic miuiary receiving less than a specified dose versus the dose (integral disease. This method remains computationally time consum or cumulative form) (Fig. 4). It therefore graphically presents

RADIATION DOSIMETRY •Zanzonico 303 the minimum, mean, and maximum doses and the dispersion 15—25rad to the fetal total body. Not surprisingly, with about the mean dose. The greater this dispersion, the more radiogenic destruction of the fetal thyroid and thyroid nonuniform is the dose distribution. hormone deficiency in utero, fetal hypothyroidism and An important practical component ofmacroscopic nonuni congenital cretinism have been shown to result after radio form dosimetry is the ability to fuse, or register, tomographic images from multiple modalities (67,112—115).Dose distri butions, calculated from 3-dimensional activity distributions measured by scintigraphic imaging (i.e., SPECT or PET), may be presented as isodose contours (Fig. 4) or color-coded images (Fig. 5). By image fusion, such isodose contours or color-coded images can be superimposed on or juxtaposed with the corresponding anatomy to allow correlation of doses with tumor and normal organs (as imaged by CT or MRI) (29,60,76,114).

“SENSITIVE―POPULATiONS The administration of certain radioactive materials, even in diagnostic amounts, to certain sensitive populations, including pregnant females, nursing mothers, and children, remains a matter of concern in nuclear medicine. Pregnant Females / By modification of the standard anthropomorphic adult phantom (21,22), increasingly accurate anatomic models of the fetus and pregnant female have been developed. These include models of the pregnant female at the beginning of B 0.14 j4 —[email protected] pregnancy (representing the embryo as a small unit density @ sphere located at the uterus) (116) and at the end of the first 0.12 and third trimesters (117). Radiopharmaceutical kinetic data ] in utero and, therefore, fetal dose estimates are quite limited, 0.10 however. Published fetal absorbed-dose estimates are gener 0.08 ally 0. 1 cGy (0. 1 rad)/37 mBq (1 mCi) administered to the I mother (116,118-122). 0.06 A particularly worrisome issue is radioiodine administra tion to pregnant females (123,124). The fetal thyroid begins concentrating iodine at 12—15wk ofgestation. At 16—24wk, ‘31I-iodidedelivers a very large absorbed dose of 1500—6000 1::: cGy/37 MBq to the fetal thyroid and an absorbed dose of 0.00 3—5cGy/37 MBq to the fetal total body, depending on maternal thyroid uptake (124). For a hyperthyroid therapy AbsorbedDoee(Gy) administration of 185 MBq (5 mCi), 7,500—30,000 cGy (rad) would therefore be delivered to the fetal thyroid and C

FIGURE4. (A)Dosedistribution,calculatedbyvoxelS factor methodand displayedas color-codedisodosecontourssuperim posedonbremsstrahlungSPECTimageofliver,afterinjectionof 32@ chromic phosphate colloid (629 MBq [17 mCi]) into patient with nonresectable hepatic metastases. Isodose curves are shownfor 10% (yellow),30% (red),50% (blue),70% (orange), and 90% (green)of maximumvoxeldoseof415 Gy (41,500rad). (B)Correspondingdifferentialdose-volumehistogram.(C)Corre sponding integral dose-volume histogram. Abscissa axes of histogramsweretruncatedat 200 Gy (20,000rad) becausesuch I small fraction of liver receiveddoses in excess of this dose. As shownby dottedlines, “ideal―differentialand integraldose volume histogramwould be vertical line and rectangle, respec 75 100 125 tively, indicating that entire tumor or tissue received uniform AbsorbedDoee(Gy) singledose. (Adaptedwith permissionof (29).)

304 THE Joui@L OF NUCLEARMEDICINE •Vol. 41 •No. 2 •February 2000 @ @:

@ / __

C

S I

D 100

80

FIGURE5. (A) SeriesoftransverseCT imagesthroughliver(blueregionofinterest [AOl])of patientwithcolorectalcarcinoma E and largehepaticmetastasis(greenAOl). (B) Correspondingseries of transverse SPECTimagesafteradministrationof 1311.. > CC49anti-TAG-72murinemonoclonalanti body ( 132)show liver activitydelineatedby blue AOl and tumoractivitydelineatedby 20 green AOl. (C) Dose distributionIn hepatic metastasis,calculatedbyconvolutionof photonpoint-kernelwith3-dimensionalac 0 tivity distribution and by assuming complete [email protected] 0.02 0.04 0.06 0.08 0.1 playedas “thermal―colorcoded,from black representing0 dose to white representing maximumdoseof0.10Gy—37MBq(10 Absorbed Dose (GyImCi) rad/mCi). (D) Integral dose—volumehisto gram.(Adaptedwithpermissionof(114).)

iodine therapy of hyperthyroidism or thyroid cancer in recommended different interruption periods before resuming pregnant females. It is therefore critical to avoid radioiodine breast-feeding after administration of radiopharmaceuticals. administration to the pregnant patient, even in diagnostic Although there is no absolute consensus, the following are amounts. representative of the published recommendations: 24 h after NursingMothers any administration of @‘Tc,2—4wk after administration of 67Ga-citrate, and permanently for the current nursing infant Diagnostic radiopharmaceuticals administered to lactat after any administration of ‘@‘I(125—128). ing women can achieve high concentrations in breast milk and deliver potentially significant radiation doses to nursing Children infants (125—128).For example, the cumulative breast milk With the publication of the week by Christy and Ecker activity ranged from 0.03% to 27% ‘@‘Iadministered to 6 man (21), anthropomorphic models are now available for women for thyroid uptake studies (129). Using a variety of newborns and 1-, 5-, 10-, and 15-y-old children. Using the dosimethc criteria (e.g., an effective dose equivalent to the MIRDOSE 3.1 computer program and available kinetic data nursing infant of 0.1 cSv [10.1 rem]), several authors have in conjunction with these models, Stabin and Gelfand (130)

RADIATION Dosm@iE@ri@Y•Zanzonico 305 estimated organ doses per unit administered activity for 25. Johnson T. MABDOS: a generalized program for internal dosimeuy. Comput children of different ages for many commonly used diagnos Methods Programs Biomed. 1988;37:159—167. 26. Johnson TIC,McClure D, McCourt S. MABDOSE: part 1: Characterization of a tic radiopharmaceuticals. Assuming typical values of admin general purpose dose estimation code. MedPhys. 1999;26:l389—1395. istered activity, the doses, the effective dose equivalent, and 27. JohnsonTK, McClure D, McCourt S. MABDO5E: part 2: Validation ofa general the effective dose per procedure were tabulated. Comprehen purpose dose estimation code. MedPhys. l999;26:1396—1403. 28. Goddu S. Howell R, Bouchet L, et al. MIRD Cellular S Factors: Self-Absorbed sive age-dependent dosimetric data are also presented in Dose per Unit Cwnulated Activity for Selected Radionuclides and Monoener ICRP publication 53 (20). getic Electrons and Alpha Particle Emitters Incorporated into Different Cell Conipart,nents. Reston, VA: Society of Nuclear Medicine; 1997:183. 29. Bolch WE, Bouchet LG. Robertson J5, et al. MIRD pamphlet No. 17: the REFERENCES dosimetry of nonuniform activity distributions—radionuclide S values at the voxellevel.JNuclMed.1999;40:11S—36S. I. Stabin MG. TagessonM, ThomasSR. Ljungberg M, StrandSE. Radiation 30. Zanzonico P, Brill A, Becker D. Radiation dosimetry. In: Wagner H. Szabo Z, dosimetry in nuclear medicine. AppiRadiat Iso:. i999;50:73-87. Buchanan J, eds. Principles ofNuclearMedicine. 2nd ed Philadelphia, PA: WB 2. Howell RW. Wessels BW, Loevinger R, et al. The MIRI) perspective 1999. J SaundersCo.; 1995:106-134. Nuci Med. 1999;40:3S—iOS. 31. Werner S. Radioiodine. In: Werner S. Ingbar S. ads. The Thyroid. 4th ed. 3. FisherDR.Radiationdosimetryforradioimmunotherapy:anoverviewofcurrent Hagerstown, MD: Harper & Row; 1978:827—835. capabilities and limitations. Cancer. 1994;73:905—911. 32. Marinelli L Quimby E, Hine G. Dosage determination with radioactive isotopes: 4. GoodheadD. Relationshipof microdosimetrictechniquesto applicationsin part 2. Practical considerations in therapy and protection. AiR. 1948;59:260- biological systems. In: Kase K, Bjarngard B, Attix F, ads. The Dosimetry of 280. Ionizing Radiation. Orlando, FL: Academic Press; 1987:1—89. 33. Marinelli L, Quimby E, Hine G. Dosage determination with radioactive isotopes: 5. Rossi H. Microscopic energy distributions. In: Attix F, RoeschW,Tochiin E, eds. part I. Fundamental dosage formulae. Nucleonics. 1948;2:56-66. Radiation Dosimetry. New York, NY: Academic Press; 1968:43-92. 34. Marinelli L, Quimby E, Hine G. Dosagedetermination with radioactive isotopes: 6. International Commission on Radiation Units and Measurements. Mkrodosim part 2. Biological considerations and practical applications. Nucleonics. 1948;2: etry. ICRU report no. 36. Bethesda, MD: international Commission on Radiation 44—49. Units and Measurements; 1983. 35. Becker D, Hurley J. Radioiodine treatment of hyperthyroidism. In: Sandier M. 7. Fisher D. The microdosimetry of monoclonal antibodies labeled with alpha Patton J, Coleman R, GottschalkA, Wackers F, Hoffer P, eds. Diagnostic Nuclear particles. In: Schialke-Stelson A, Watson E, eds. Proceedings of the Fourth Medicine. 3rd ad. Baltimore, MD: Williams & Wilkins; 1996:943-958. Infrrnational Radiopharmaceutical Dosimetry Symposiwn. Oak Ridge. TN: 36. Lazarus JH. Guidelines for the use of radioiodine in the management of United States Depamnent of Energy and Oak Ridge Associated Universities; hyperthyroidism: a summary. Prepared by the Radioiodine Audit Subcommittee 1986:26—36. of the Royal College of Physicians Committee on Diabetes and Endocrinology 8. Humm J. A microdosimetric model of astatine-211 labeled antibodies for and the Research Unit of the Royal College of Physicians. J R Coil Physicians radioiminunotherapy. In: JRG4iOJ Oncol BiOLPhys. 1987;13: 1767—1773. Land.1995;29:464-469. 9. Humm JL, Roeske JC, Fisher DR. Chen GT. Microdosimetric concepts in radioimmunotherapy.MedPhys.1993;20:535—541. 37. Barandes M, Hurley J, Becker D. Implications of rapid intrathyroidal turnover 10.KellererA, ChmelevskyD. Criteriafor theapplicabilityof LET.Radio:Res. for 131-1therapy: the small pool syndrome [abstr@t]. JNuclMed. 1973;14(suppl): 379. 1975;63:226-234. 11. National Council on Radiation Protection and Measurements.SI Units in 38. DeNardo Si, DeNardO GL, O'Grady LF, et al. Treatment ofa patient with B-cell lymphoma by I-i31 LYM-1 monoclonal antibodies. mt i Biol Markers. Radiation Protection and Measurements. NCRP report no. 82. Bethesda, MD: National Councilon Radiation Protection and Measurements; 1985. 1987;2:49—53. 12. International Commission on Radiological Protection. Recommendations of the 39.DeNardoDA,DeNardoGL,YuanA, etal.Predictionof radiationdosesfrom International Commission on Radiological Protection. ICRP publication no. 60. therapy using tracer studies with iodine-l31-Iabeled antibodies. J Nucl Med. Elmsford, NY: international Commission on Radiological Protection; 1977. l996;37:1970—1975. 13. National Council on Radiation Protection and Measurements. Limitations of 40. SgourosG. Treatmentplanningfor internalemittertherapy:methods,applica Esposure to Ionizing Radiation. NCRP repoil no. 91. Bethesda, MD: National tions and clinical implications. in: Schlafke-Stelson A, Stabin M, Sparks R, eds. Council on Radiation Protection and Measurements; 1987. Proceedings ofthe Sixth International Radiopharmaceutical Dosimetry Sympo 14. Zanzomco P. The fallacy of the chest x-ray as a basis for comparing radiogemc sium. ORISE 99-0164. Oak Ridge, TN: United States Department of Energy risks [abstract]. JNucIMed. 1993;34(suppl):133P. and Oak RidgeAssociated Universities; 1999:13—25. 15. international Commission on Radiological Protection. Recommendations of the 41. Zanzonico P, Edwards C, Sgouros G, et al. Practical dosimetry: quantitative International Commission on Radiological Protection. ICRP publication no. 60. imaging in radionudide therapy. In: Adelstein S, Kassis A, Burt R, eds. Elmsford, NY: International Commission on Radiological Protection; 1991. Dosimetry ofAdministered Radionuclides. Washington. DC: American College 16. National Council on Radiation Protection and Measurements. limitations of ofNuclear Physicians/United States Department ofEnergy; 1989:275—294. Erposure to Ionizing Radiation. NCRP report no. 116. Bethesda, MD: National 42. Erdi AK, Erdi YE, Yorke ED. Weasels BW. Treatment planning for radio Councilon Radiation Protection and Measurements; 1993. immunotherapy. Phys MedBiol. 1996;41:2009—2026. 17.WatsonE.Foreword.JNucIMed.1999;40:15-2S. 43. Ott Ri. Imaging technologies for radionuclide dosimetry. Phys Med Biol. 18.LoevingerR.BudingerT,WatsonE,et al.MIRDPrimerforAbsorbedDose 1996;41:l885—1894. Cakulations. New York, NY: Society ofNuclear Medicine; 1991. 44. SiegelJA, Thomas SR. Stubbs lB. et al. MIRD Pamphlet No. 16.Techniques for 19. LoevingerR. The MIRD perspective.In: AdelsteinS. KassisA, Burt R, eds. quantitative radiopharinaceutical biodistribution data acquisition and analysis Dosinieuy ofAdininistered Radionuclides. Washington, DC: American College for use in human radiation dose estimates. I NuclMed. 1999;40:37S—6iS. of Nuclear PhysicianslUnited States Department of Energy; 1989:29—43. 45. Benua R, Cicale NR, Sonenberg M. The relation of radioiodine dosimetry to 20. international Commission on Radiological Protection. Radiation Dose to results and complications in the treatment of metastatic thyroid cancer. AiR. Patients from Rdiopham@aceuticaLs@ ICRP publication no. 53. Oxford, UK: 1962;87:l71—182. international Commission on Radiological Protection; 1983. 46. Furhang EE, Larson SM, Buranapong P. Humm JL. Thyroid cancer dosimetry 21. Cristy M, Eckennan K. Specific Absorbed Fractions ofEnergy at Various Ages using clearance fitting. JNuclMed. 1999;40:13l—136. from Internal Photon Sources (I.VII). Oak Ridge National Laboratory Report 47. Kaminski MS. Zasadny KR, Francis IR. et al. Radioimmunotherapy of B-cell ORNIJFM-8381/Vi-7. Springfield, VA: National Technical Information Ser lymphomawith['311]anti-Bl (anti-CD2O)antibody.NEnglJMed. 1993;329:459— vice, United States Department ofCommerce; 1987. 465. 22. SnyderW, Ford M, WarnerG. Estimates ofabsorbed fractions for monoenergetic 48. KaminskiMS.zasadnyKR, FrancisIR, etal. lodine-131-anti-B1radioimmuno photon sources uniformly distributed in various organs of a heterogeneous therapy for B-cell lymphoma. I Clin Oncol. 1996;14:i974—1981. phantom: MIRD pamphlet no. 5. JNuclMed. i969;10(suppl 3):5—52. 49. WahI RL, Zasadny KR, MacFarlane D, et al Iodine-131 anti-Bl antibody for 23. Stabin MG. MIRDOSE: personal computer software for internal dose assess B-cell lymphoma: an update on the Michigan phase I experience. J NucI Med. ment in nuclear medicine. JNucIMed. 1996;37:538—546. i998;39(suppl):215—27S. 24. Sgouros G, Bigler R, Zanzonico P. DOSCAL: a tumor-incorporating mean 50. Zasadny K, Gates V. Fisher S. Kaminski M, Wahi R. Correlation of dosimetric absorbed dose calculation program [abstract]. I Nuci Med. l988;29(suppl):874. parameters with hematological toxicity after radioimmunotherapy of non

306 THE Joui@i. OF NUCLEARMEDICINE •Vol. 41 •No. 2 •February 2000 Hodgkin's lymphoma with 1-131anti-Bl: utility ofa new parameter—Totalbody 77. Liu A, Williams L, Wong J, Raubitschek A. A voxel source kernel (VSK) method dose-lean [abstract]. JNuclMed. 1995;36(suppl):214P. for rapid, patient-specific dose estimates in radioimmunotherapy (RIT) [ab 51. Liu A, Williams LE, Demidecki AJ, Raubitschek AA, Wong JY. Modeling of stract]. JNuclMed. 197;38(suppl):106P. tumor uptake to determine the time-dose-fractionation effect in radioimmuno 78. Adelstein 5, Kassis A, Sastry K. Cellular vs. organ approaches to dose estimates. therapy [letter]. J Nuci Med. l994;35: 1561—1564. In: Schlafke-Stelson A, Watson E, ads. Proceedings ofthe Fourth International 52. Odom-Maiyon U, Williams LE, Chai A, et al. Pharmacokinetic modeling and Radiopharmaceutical Dosimetry Symposium. Oak Ridge. TN: United States absorbed dose estimation for chimeric anti-CEA antibody in humans. J Nucl Department ofEnergy and Oak RidgeAssociated Universities; 1986:13—25. Med. 1997;38:1959—1966. 79. GodduSM, Rio DV, Howell RW.Multicellular dosimetryfor micrometastases: 53. Sgouros G. Plasmapheresis in radioimmunotherapy of micrometastases: a dependence of self-dose versus cross-dose to cell nuclei on type and energy of mathematical modeling and dosimetrical analysis. J Nucl Med. 1992;33:2167— radiation and subcellulardistribution ofradionuclides. JNuclMed. 1994;35:521— 2179. 530. 54. Sgouros G, Graham MC, Divgi CR, Larson SM, Scheinberg DA. Modeling and 80. Griffiths GL, Govindan SV, Sgouros G. Ong GL Goldenberg DM, Mattes Mi. dosimetry of monoclonal antibody M195 (anti-CD33) in acute myelogenous Cytotoxicity with Auger electron-emitting radionuclides delivered by antibodies. leukemia. I NucIMed. l993;34:422—430. IntJ Cancer. 1999;8l:985—992. 55. Strand SE, Zanzonico P. Johnson TK. Pharmacokinetic modeling. Med Phys. 81. Humm IL, Howell RW, Rao DV. Dosimetry ofAuger-electron-emitting radionu 1993;20:515—527. clides: report no. 3 of AAPM Nuclear Medicine Task Group No. 6. Med Phys. 56. Willins JD, Sgouros G. Modeling analysis of platinum-l95m for targeting l994;21 :1901—1915.[Published correction appears in Med Phys. 1995;22: 1837.] individual blood-borne cells in adjuvant radioimmunotherapy. J Nucl Med. 82. Humm JL, Macklis RM, Lu XQ, et al. The spatial accuracy of cellular dose 1995;36:315—319. estimates obtained from 3D reconstructed serial tissue autoradiographs. Phys 57. Zanzonico P, Bigler R, Primus F, et al. A compartmental modeling approach to MedBiol. 1995;40:163—180. the radiation dosimetry ofradiolabeled antibody. In: Schlafke-StelsonA, Watson 83. Macklis RM, Lin JY, BeresfordB, Atcher RW,Hinesii. Humm JL. Cellular E, eds. Proceedings ofthe Fourth International Radiopharmaceutical Dosimetry kinetics, dosimetry, and radiobiology of alpha-particle radioimmunotherapy: Symposium. Oak Ridge, TN: United States Department of Energy and Oak induction of apoptosis. Radiat Res. 1992;l30:220—226. Ridge Associated Universities; 1986:421—445. 84. O'DonoghueJA. Wheldon TE. Targetedradiotherapyusing Auger electron 58. Furhang EE, Chui CS, Sgouros G. A Monte Carlo approach to patient-specific emitters. Phys Med Biol. l996;41 :1973-1992. dosimetry. MedPhys. l996;23:l523—1529. 85. O'DonoghueJA. Strategiesfor selectivetargetingof Augerelectronemittersto 59. FurhangEE.ChuiCS,KolbertKS,LarsonSM, SgourosG. Implementationof a tumor cells. JNucl Med. 1996;37(suppl):3S—6S. Monte Carlo dosimetry method for patient-specific internal emitter therapy. Med 86. Wrenn ME, Howells GP, Hairr LM. Paschoa AS. Auger electron dosimetry. Phys. 1997;24: 1163—1172. Health Phys. l973;24:645—653. 60. Erdi AK, Wessels BW. Deiager R, et al. Thmor activity confirmation and isodose 87. Yorke ED, Williams LE, Demidecki AJ, Heidorn DB, Roberson PL, Weasels curve display for patients receiving iodine-13l-labeled 16.88 human monoclonal BW. Multicellular dosimetry for beta-emitting radionuclides: autoradiography, antibody. Cancer. l994;73:932—944. thermoluminescent dosimetry and three-dimensional dose calculations. Med 61. Erdi AK, Yorke ED, Loew MH, Erdi YE, Sarfaraz M, Wessels BW. Use of the Phys. l993;20:543—550. fast Hartley transform for three-dimensional dose calculation in radionuclide 88. Zalutsky MR, Stabin MG, LarsenRH, Bigner DD. Tissuedistribution and therapy. Med Phys. l998;25:2226—2233. radiation dosimetry of astatine-211-labeled chimeric 8lC6, an alpha-particle 62. Giap HB. Macey Di. Bayouth JE, Boyer AL. Validation of a dose-point kernel emitting immunoconjugate. NuclMedBiol. 1997:24:255-261. convolution technique for internal dosimetry. Phys Med Biol. 1995;40:365—38l. 89. Kassis A!, Adelstein SJ, Haydock C, Sastry KS. Radiotoxicity of “Seand @@S: 63. Giap HB, Macey DJ, Podoloff DA. Development of a SPEC'I'-based three theory and application to a cellular model. RadiatRes. 1980;84:407—425. dimensional treatment planning system for radioimmunotherapy. J Nucl Med. 90. KassisA!, AdelsteinSJ, HaydockC, SastryKS, McElvany KD. Welch Mi. 1995;36:1885—1894. Lethality of Auger electrons from the decay of bromine-77 in the DNA of 64. HowellRW,RaoDV,SastryKS.Macroscopicdosimetryfor radioimmuno mammalian cells. RadiatRes. 1982;90:362—373. therapy: nonuniform activity distributions in solid tumors. Med Phys. 1989;l6: 91. Kassis A!, Adelstein SJ, Haydock C, Sastry KS. Thalhium-201: an experimental 66—74. and a theoretical radiobiological approach to dosimetry. J Nucl Med. l983;24: 65. Humm IL. Dosimetric aspects of radiolabeled antibodies for tumor therapy. J 1164—1175. NucIMed. 1986;27:1490—1497. 92. KassisAl, SastryKS,AdelsteinSi. Intracellulardistributionandradiotoxicityof 66. Humin JL, Cobb LM. Nonuniformity of tumor dose in radioimmunotherapy. J chromiwn-51 in mammalian cells: Auger-electron dosimetry. J Nucl Med. Nucl Med. 1990:31:75—83. 1985;26:59—67. 67. Koral KF. Zasadny KR, Kessler ML, et al. CT-SPECT fusion plus conjugate 93. Kassis A!, Fayad F, Kinsey BM. Sastry KS, Taube RA, Adelstein Si. views for determining dosimetry in iodine-l3l-monoclonal antibody therapy of Radiotoxicity of ‘@‘lin mammalian cells. Radiat Res. 1987;lll :305-3 18. lymphoma patients. J NuclMed. 1994;35: 1714—i720. 94. KassisAl, SastryKS,AdelsteinSJ.Kineticsof uptake,retention.andradiotoxic 68. Kwok CS, PrestwichWV, Wilson BC. Calculation of radiation dosesfor ity of ‘@IUdRin mammalian cells: implications of localized energy deposition nonuniformity distributed beta and gamma radionuclides in soft tissue. Med by Auger processes. Radiat Res. 1987;109:78—89. Phys.1985:12:405—412. 95. KassisA!. The MIRD approach:rememberingthe limitations. J NucI Med. 69. Leichner PK, Kwok CS. Tumor dosimetry in radioimmunotherapy: methods of 1992;33:781—782. calculation for beta particles. Med Phys. 1993;20:529—534. 96. Makrigiorgos GM, Adelstein Si, Kassis Al. Limitations of conventional internal 70. Liu A, Williams LE, Lopatin G, Yamauchi DM, Wong JY. Raubitschek AA. A dosimetry at the cellular level. iNucI Med. 1989;30:l856—1864. radionuclide therapy treatment planning and dose estimation system. J NucI 97. Siegel J, Weasels B, Watson E, et al. Bone marrow dosimetry and toxicity in Med. 1999;40:115l—1153. radioimmunotherapy. Antibody Immunocon Radiopharm. 1990:3:213-223. 7 1. Ljungberg M, Strand SE. Dose planning with SPECT. tnt J Cancer Suppi. 98. DeNardoGL, DeNardo Si, Macey Di, Shen S. Kroger LA. Overview of 1988;2:67—70. radiation myelotoxicity secondary to radioimmunotherapy using ‘311-Lym-las a 72. Mayer R, Dillehay LE, Shao Y, et al. A new method for determining dose rate model. Cancer. 1994;73:1038-1048. distribution from radioimmunotherapy using radiochromic media Int I Radiat 99. Bigler R, ZanzonicoP.PrimusF,et al. Bonemarrowdosimetryfor monoclonal Oncol Biol Phys. 1994;28:505—5I3. antibody therapy. In: Schlafke-Stelson A, Watson E, eds. Proceedings of the 73. Sgouros G, Chiu 5, Pentlow KS, et al. Three-dimensional dosimetry for Fourth International Radiopharmaceutical Dosimetry Symposium. Oak Ridge, radioimmunotherapy treatment planning. iNucIMed. 1993;34:1595—1601. TN: United States Department of Energy and Oak Ridge Associated Universi 74. Uchida I, Yamada Y, Oyamada H, Nomura E. Calculation algorithm of ties; 1986:535—544. three-dimensional absorbed dose distribution due to in vivo administration of 100. Sgouros G. Bone marrow dosimetry for radioimmunotherapy: theoretical nuclides for radiotherapy [in Japanese]. Kaku Igaku. 1992;29: 1299—1306. considerations. JNuclMed. l993;34:689—694. 75. Wessels BW. Yorke ED, Bradley EW. Dosimetry of heterogeneous uptake of 101. Macey Di, DeNardo SJ, DeNardO GL, DeNardo DA, Shen S. Estimation of radiolabeled antibody for radioimmunotherapy. Front Radiat Ther Oncol. radiation absorbed doses to the red marrow in radioimmunotherapy. Clin Nucl l990;24: 104—108. Med. 1995:20:117—125. 76. Akabani G. Hawkins W, Eckblade M, Leichner P. Patient-specific dosimetry 102. DeNardo DA, DeNardo GL, O'Donnell RT. et al. Imaging for improved using quantitative SPECT imaging and three-dimensional discrete Fourier prediction of myelotoxicity after radioimmunotherapy. Cancer. 1997:80:2558— transform convolution. J Nuci Med. l997;38:308—314. 2566.

RADIATIONDOSIMETRY•Zanzonico 307 103. Lim S. DeNardo GL, DeNardo DA, O'Donnell RT, Yuan A, DeNardo Si. International Radiopharmaceutical Dosimetry Symposium. Oak Ridge, TN: Prediction of myelotoxicity using semi-quantitative marrow image scores. I United States Department of Energy and Oak Ridge Associated Universities; NuclMed. 1997;38:1749—1753. 1986:89—104. 104. Lim SM, DeNardo GL, DeNardo DA, et al. Prediction of myelotoxicity using I18. Hedrick WR, DiSimone RN, Wolf BH, Langer A. Absorbed dose to the fetus radiation doses to marrow from body, blood and marrow sources. I Nucl Med. during bone scintigraphy. Radiology. 1988;168:245—248. l997;38:1374—l378. 119. Cloutier Ri, Smith SA, Watson EE, Snyder WS, Warner GG. Dose to the fetus 105. Zanzonico P. Sgouros G. Predicting myelotoxicity in radioimmunotherapy: what from radionucides in the bladder. Health Phys. 1973;25: 147—161. does dosimetry contribute? JNuclMed. 1997;38:1753—1754. 120. Husak V, Wiedermann M. Radiation absorbed dose estimates to the embryo from 106. Sgouros G, Divgi C, Scott A, Williams J, Larson S. Hematologic toxicity in some nuclear medicine procedures. Eur J NucI Med. 1980:5:205—207. radioimmunotberapy: an evaluation ofdifferent predictive measures [abstract]. I 121. Marcus CS, Mason GR, Kuperus JH, Mena I. Pulmonary imaging in pregnancy: Nucl Med. 1996;37(suppl):43P.-44P. maternal risk and fetal dosimetry. Clin NuclMed. 1985;10:l—4. 107. Sgouros G, Deland D, Loh A, Divgi C, Larson S. Marrow and whole-body 122. Ponto JA. Fetal dosimetry from pulmonary imaging in pregnancy: revised absorbed dose vs. marrow toxicity following ‘@‘I-G250antibody therapy in estimates. Clin NuclMed. l986;ll:l08—l09. patients with renal cell carcionoma[abstract]. iNucIMed. 1997;38(suppl):252P. 123. Stabin MG, Watson EE, Marcus CS, Salk RD. Radiation dosimetry for the adult 108. O'Donoghue JA. Implications of nonuniform tumor doses for radioimmuno female and fetus from iodine-131 administration in hyperthyroidism. I NucI therapy. JNuclMed. 1999;40:1337—134l. Med. 1991;32:808—8l3. 109. Human IL, Macklis RM, Bump K, Cobb LM, Chin LM. Internal dosimetry using 124. Zanzonico P.Becker D. Radiation hazards in children born to mothers exposed to data derived from autoradiographs. I NuclMed. l993;34: 1811—1817. ‘@‘I.In: Beckers C, Reinwein D, eds. The Thyroid and Pregnancy. Stuttgart, 110. Sgouros G, Barest G, Thekkumthala J, et al. Treatment planning for internal Germany: Schattauer, 1991 :I89—202. radionucide therapy: three-dimensional dosimetry for nonuniformly distributed 125. Ahlgren L, Ivarsson 5, Johansson L, Mattsson 5, Nosslin B. Excretion of radionucides.JNuclMed. 1990;3l:1884—1891. radionucides in human breast milk after the administration of radiopharmaceuti Ill. Tagesson M, Ljungberg M, Strand SE. A Monte-Carlo program converting cals. JNuclMed. 1985;26:l085—1090. activity distributions to absorbed dose distributions in a radionuclide treatment 126. Mountford PJ, Coakley Ai. A review of the secretion of radioactivity in human planning system. Acta Oncol. l996;35:367—372. breast milk: data, quantitative analysis and recommendations. NucI Med 112. ScottAM, Macapinlac H, Zhang JJ, et al. Clinical applications offusion imaging Commun. 1989;lO: 15—27. in oncology. NucI Med Biol. 1994;21 :775—784. 127. Romney B, Nickoloff E, Esser P. Alderson P. Radionuclide administration to I 13. Scott AM, Macapinlac H, Thang J. et al. Image registration of SPECI' and CT nursing mothers: mathematically derived guidelines. Radiology. l986;l60:549— images using an external fiduciary band and three-dimensional surface fitting in 554. metastatic thyroid cancer. INuciMed. 1995;36:l00-l03. 128. Rubow 5, Klopper J, Wasserman H, Baard B, van Niekerk M. The excretion of 114. Kolbert KS, Sgouros G, Scott AM, et al. Implementation and evaluation of radiopharmaceuticals in human breast milk: additional data and dosimetry. EurJ patient-specific three-dimensional internal dosimetry. INuciMed. 1997;38:301— NuciMed. l994;21:144—153. 308. 129. Weaver J, Kamm M, Dobson R. Excretion of radioiodine in human milk. JAMA. 115. Koral I(F, Lin 5, Fessler JA, Kaminski MS. WahI RI. Preliminary results from 1960:872—875. intensity-based CT-SPECTfusion in 1-131 anti-Bl monoclonal-antibody therapy 130. Stabin MG, Gelfand Mi. Dosimetry of pediatric nuclear medicine procedures. Q oflymphoma. Cancer. 1997;80:2538—2.544. JNuclMed. l998;42:93—112. 116. Smith EM. Warner GO. Estimates of radiation dose to the embryo from nuclear 131. Sgouros G, Jureidini IM. Scott AM, Graham MC, Larson SM, Scheinberg DA. medicine procedures. JNucl Med. l976;17:836—839. Bone marrow dosimetry: regional variability of marrow-localizing antibody. I 117. Watson E, Stabin M, Eckerman K, Cristy M, Ryman J. Davis J. Model 2: Nucl Med. 1996;37:695—698. mathematical models of the pregnant woman at the end of the first and third 132. Divgi CR, Scott AM, Dantis L et xi. Phase I radioimmunotherapy trial with trimesters. In: Schlafke-Stelson A, Watson E, eds. Proceedings of the Fourth iodine-l31-CC49 in metastatic colon carcinoma JNuclMed. l995;36:586—592.

308 THE JOURNALOF NUCLEARMEDICINE •Vol. 41 •No. 2 •February 2000