permanent prostate seed implant brachytherapy: report of ... · association of physicists in...

Permanent prostate seed implant brachytherapy: Report of the AmericanAssociation of Physicists in Medicine Task Group No. 64 a…

Yan Yub)

Department of Radiation Oncology, University of Rochester, Rochester, New York 14642

Lowell L. AndersonDepartment of Medical Physics, Memorial Sloan-Kettering Cancer Center, New York, New York 10021

Zuofeng LiDepartment of Radiation Oncology, University of Florida, Gainesville, Florida 32610

David E. MellenbergRadiation Oncology Division, University of Iowa, Iowa City, Iowa 52242

Ravinder NathDepartment of Therapeutic Radiology, Yale University School of Medicine, New Haven, Connecticut 06510

M. C. SchellDepartment of Radiation Oncology, University of Rochester, Rochester, New York 14642

Frank M. WatermanDepartment of Radiation Oncology, Thomas Jefferson University, Philadelphia, Pennsylvania 19107

Andrew WuDepartment of Radiation Oncology, Allegheny General Hospital, Pittsburgh, Pennsylvania 15212

John C. BlaskoUniversity of Washington Medical Center, Seattle, Washington 98195

~Received 28 April 1999; accepted for publication 20 July 1999!

There is now considerable evidence to suggest that technical innovations, 3D image-based plan-ning, template guidance, computerized dosimetry analysis and improved quality assurance practicehave converged in synergy in modern prostate brachytherapy, which promise to lead to increasedtumor control and decreased toxicity. A substantial part of the medical physicist’s contribution tothis multi-disciplinary modality has a direct impact on the factors that may singly or jointly deter-mine the treatment outcome. It is therefore of paramount importance for the medical physicscommunity to establish a uniform standard of practice for prostate brachytherapy physics, so thatthe therapeutic potential of the modality can be maximally and consistently realized in the widerhealthcare community. A recent survey in the U.S. for prostate brachytherapy revealed alarmingvariance in the pattern of practice in physics and dosimetry, particularly in regard to dose calcula-tion, seed assay and time/method of postimplant imaging. Because of the large number of start-upprograms at this time, it is essential that the roles and responsibilities of the medical physicist beclearly defined, consistent with the pivotal nature of the clinical physics component in assuring theultimate success of prostate brachytherapy. It was against this background that the RadiationTherapy Committee of the American Association of Physicists in Medicine formed Task Group No.64, which was charged~1! to review the current techniques in prostate seed implant brachytherapy,~2! to summarize the present knowledge in treatment planning, dose specification and reporting,~3!to recommend practical guidelines for the clinical medical physicist, and~4! to identify issues forfuture investigation. ©1999 American Association of Physicists in Medicine.@S0094-2405~99!00910-4#

Key words: brachytherapy, interstitial brachytherapy, prostate seed implant, quality assurance,standards of practice

TABLE OF CONTENTS

GLOSSARY OF ABBREVIATIONS . . . . . . . . . . . . . 2055I. INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . .2055II. REVIEW OF CURRENT TECHNIQUES. . . . . . . . 2057

A. Overview of contemporary techniques. .. . . . . . 2057B. Treatment planning techniques. . . . . . . . . . . . . . 2057

2054 Med. Phys. 26 „10…, October 1999 0094-2405/99/26 „1

1. Ultrasound volume study. . . . . . . . . . . . . . . . 20582. Pubic arch. . . . . . . . . . . . . . . . . . . . . . . . . . . .20583. Seed distribution. . . . . . . . . . . . . . . . . . . . . . .20584. Urethra. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20585. Rectum. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20596. Dose margin. . . . . . . . . . . . . . . . . . . . . . . . . .20597. Intraoperative planning. . . . . . . . . . . . . . . . . . 2059

20540…/2054/23/$15.00 © 1999 Am. Assoc. Phys. Med.

1

et

eaf

d-

n-s

olt

ds

-

d

od-

--e

2055 Yu et al. : Task Group No. 64 2055

C. Equipment and applicators. . . . . . . . . . . . . . . . . . 2060D. Source type, assay and preparation. . . . . . . . . . . 2060E. Implantation procedure. . . . . . . . . . . . . . . . . . . . .2062F. Postimplant dosimetry. . . . . . . . . . . . . . . . . . . . .2063

1. Rationale. .. . . . . . . . . . . . . . . . . . . . . . . . . . .20632. Technical issues. . . . . . . . . . . . . . . . . . . . . . .2063

III. REVIEW OF DOSIMETRIC ASPECTS. .. . . . . . 2064A. Historical perspective on dosimetry. . . . . . . . . . 2065B. Dosimetry data revision. . . . . . . . . . . . . . . . . . . . 2066C. Dose specification and reporting. . . . . . . . . . . . . 2066D. Dosimetric uncertainties. . . . . . . . . . . . . . . . . . . . 2067E. Treatment plan evaluation. . . . . . . . . . . . . . . . . . 2067

1. SmPD. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20672. Dose uniformity. . . . . . . . . . . . . . . . . . . . . . .20683. Dose conformity.. . . . . . . . . . . . . . . . . . . . . .20684. Dose-volume histogram. . . . . . . . . . . . . . . . . 2068

F. Treatment plan optimization. . . . . . . . . . . . . . . . 2068IV. CURRENT RECOMMENDATIONS.. . . . . . . . . . 2069

A. Equipment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2069B. New radionuclide designs. . . . . . . . . . . . . . . . . . 2069C. Seed assay. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2070

GLOSSARY OF ABBREVIATIONS

ADCL Accredited Dosimetry CalibrationLaboratory.

D100,D90,D80 Dose to 100%, 90%, 80% of the targvolume for dosimetric evaluation.

DVH Dose-volume histogram.Gleason score A pathological grading system for m

suring the degree of differentiation oprostate tumors.

MPD Matched peripheral dose, typically usein conjunction with the ellipsoidal approximation for the prostate volume.

mPD Minimum peripheral dose. Used idosimetric planning, the mPD corresponds to the isodose surface that juencompasses the planning target vume. This report recognizes tha

ligee.

miag

tenop

Medical Physics, Vol. 26, No. 10, October 1999

D. Changes to the dose value due to TG43. . . . . . . 2070E. Dosimetric planning. . . . . . . . . . . . . . . . . . . . . . .2070F. Implantation procedure. . . . . . . . . . . . . . . . . . . . .2070G. Patient release. . . . . . . . . . . . . . . . . . . . . . . . . . . .2070H. Postimplant analysis. . . . . . . . . . . . . . . . . . . . . . .2070I. Training requirement for physics personnel. . . . 2071J. Recommendations regarding commercial

treatment planning systems. . . . . . . . . . . . . . . . . 2071V. ISSUES FOR FUTURE CONSIDERATION. . . . . 207

A. Anisotropic dose calculation. .. . . . . . . . . . . . . . 2071B. Interseed effect. . . . . . . . . . . . . . . . . . . . . . . . . . .2071C. Tissue heterogeneity. . . . . . . . . . . . . . . . . . . . . . .2072D. Biological models. . . . . . . . . . . . . . . . . . . . . . . . .2072E. Relative biological effectiveness. . . . . . . . . . . . . 2072F. Time course of target volume change. .. . . . . . . 2072G. Differential dose planning and delivery. . . . . . . 2072H. Intraoperative seed localization and

dosimetry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2073I. Correlation of dosimetric and clinical

outcomes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2073ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . . . .2073

-

t-

whereas the mPD usually corresponto a prescribed dose,D100 as determinedfrom postimplant dosimetry may be significantly different from the prescribeddose.

NIST National Institute of Standards anTechnology.

PSA Prostate specific antigen.PTV Planning target volume.SmPD Total source strength required t

achieve 1 Gy in the mPD, typically usein dosimetric planning and optimization.

TRUS Transrectal ultrasound.V200,V100,V90,V80 ~fractional! Volume of the prostate tar

get for dosimetric evaluation that received 200%, 100%, 90%, 80% of thprescribed mPD.

deria-tionit is

to3, ittateseed

hy-m-s in

I. INTRODUCTION

Adenocarcinoma of the prostate is the most common manancy in man in the United States, excluding skin cancThe American Cancer Society estimated that 184,500 ncases of prostate cancer would be diagnosed in the U.S1998. Growing emphasis on prostatic specific antigen~PSA!based early detection and changes in the population degraphics in the U.S. suggest that the number of newly dnosed cases would continue to increase each year. Intreatment of prostate cancer, there is now a broad resurgof interest in the role of permanent interstitial implantationradioactive seeds. Prostate seed implants are currently

-r.win

o--

hece

fer-

formed using iodine-125 and palladium-103 sources unimaging and template guidance to deliver localized irradtion to high doses. For selected patients, seed implantaalone offers a complete course of treatment; for others,used in conjunction with external beam radiation therapythe pelvis. Based on PSA screening data in 1991 to 199was estimated that up to 10% of all newly diagnosed proscancer patients would be considered ideal candidates forimplantation as definitive radiotherapy management.1

The techniques for permanent interstitial prostate bractherapy evolved in two distinct eras. Historically, seed iplantation was performed by free-hand placement of seed

oltaedlaervehe

unf

tioti

l a

ffu

ndo-ctrceal

iveuricrais

enti

bi

arsui

tdn

thifia

necheell

ctoRein

f

e-

omyrghev-be-sy,ingionros-e ofent-d

lowdated

ug-ing,im-

ergyto

tiali-atisicsforten-

al-

t ofhy-alsedceIntetri-eseim-yet

pyla-n-theost

ant.ingrd-s in

ari-ar-

me/ber


an open surgical procedure via the retropubic approach.2 Do-simetric planning was limited to the use of nomographs flowing intraoperative measurement of the size of the prosgland.3–5 The total activity to be implanted was determinusing the average dimensions of the prostate. Postimpdosimetry was analyzed in terms of the matched periphdose~MPD!, defined as the isodose surface that would coa spatial volume numerically equal to the volume of tprostate inferred from the ellipsoidal approximation.6 Over-all, the open surgical technique suffered from substantialcertainties in dosimetric planning, implant execution adose evaluation. In contrast, contemporary techniquesprostate brachytherapy rely on three-dimensional~3D!image-based treatment planning and real-time visualizaof needle insertion and/or seed deposition. Seed implantais performed under template guidance via a transperineaproach in a percutaneous procedure typically performedan outpatient surgical setting. Holmet al.7 first described theuse of transrectal ultrasound~TRUS! for precise guidance otransperineal seed insertion in 1983. The technique wasther popularized by Blasko, Grimm, Ragde aco-workers,8–10 and has evolved into the most popular mdality for prostate seed implantation to date. Characteristithe technique is the use of TRUS for preoperative dosimeplanning and intraoperative visualization of needle plament. A somewhat different technique, developed by Wner et al.,11,12 uses computerized tomography~CT! to iden-tify the target volume for treatment planning; intraoperatneedle placement is verified under fluoroscopy using thethra as the primary landmark. Compared to the open surgtechnique, these contemporary techniques place consideemphasis on 3D conformal dosimetric planning and precplacement of the planned seed configuration in the patiGreater emphasis is also placed on careful patient selecbased on serum PSA levels and Gleason scores.

The clinical experience associated with the retroputechnique has been a subject of active investigation.13–21Theclinical results of long term studies with 10–15 yefollow-up have been mixed, partly because the techniqueseed implantation and hence the implant qualities were qvaried. In particular, Zelefsky and Whitmore20 recently re-ported the final assessment of the 15-year outcome ofhistorical series of retropubic freehand implants performethe Memorial Sloan-Kettering Cancer Center. They cocluded that the technique was associated with a greaterexpected incidence of local relapse at 15 years, and identsuboptimal dose distribution due to technical limitationsthe possible cause of the unfavorable outcome. Nathet al.21

examined the 3D dose distribution of 110 prostate implaperformed at Yale–New Haven Hospital in this era. Thidentified a number of dosimetric quality indicators to whistatistically significant differences in local recurrence-frsurvival could be attributed. Patients in the dosimetricafavorable group had 10-year survival rates higher by a faof up to 2 compared to those in the unfavorable group.view of the contemporary transperineal experience ustemplate and image guidance is still ongoing.22–33 Whilesome early studies22,31 indicate significant proportions o


-te

ntalr

n-dor

nonp-in

r-

ofic-

l-

e-albleet.

on

c

ofte

heat-aneds

tsy

yr-g

tumor-positive biopsies postimplantation and/or distant mtastases, most of the large published series24–26,29 haveshown PSA-based control rates comparable to prostatector external beam radiation. A notable study by Vijverbeet al.33 examined biopsy findings postimplantation and tquality of the implant in terms of the minimum dose deliered to the prostate. They reported significant correlationtween the implant quality and the resulting negative biopand between the implant quality and the serum PSA durfollow-up. One of the major advantages of seed implantathas been the lower morbidity rates compared to radical ptatectomy and external beam radiation therapy. The uscontemporary techniques has further reduced treatmrelated morbidities. Urinary or rectal complication ansexual dysfunction are generally reported to be relativein many recent studies.34–39 Careful treatment planning anexecution are expected to further reduce treatment-relmorbidities.

In summary, there is now considerable evidence to sgest that technical innovations, 3D image-based planntemplate guidance, computerized dosimetry analysis andproved quality assurance practice have converged in synin modern prostate brachytherapy, which promise to leadincreased tumor control and decreased toxicity. A substanpart of the medical physicist’s contribution to this multdisciplinary modality has a direct impact on the factors thmay singly or jointly determine the treatment outcome. Ittherefore of paramount importance for the medical physcommunity to establish a uniform standard of practiceprostate brachytherapy physics, so that the therapeutic potial of the modality can be maximally and consistently reized in the wider healthcare community.

Prostate seed implantation is the permanent placemenradioactive seeds in the prostate using interstitial bractherapy techniques. However, it differs from traditionbrachytherapy in three important aspects: 3D anatomy-badosimetric planning, real-time diagnostic imaging guidanand fast dose fall-off due to lower energy radionuclides.addition, it differs from remote-controlled high dose rabrachytherapy in that the radioactive source strength disbution is less amenable to optimization and alteration. Thconsiderations lead to the unique nature of prostate seedplant, which may be characterized as precision-orienteddosimetrically sensitive. To the majority of brachytherapractitioners, image-guided interstitial implantation is a retively new treatment technique. A typical implant team cosists of the radiation oncologist, the medical physicist,urologist and/or the ultrasound radiologist. At present, mof the practitioners acquire technical proficiency throughshort training course followed by actual patient treatmeThere is as yet no uniform requirement either in the traincurriculum or within the medical physics profession regaing adequate understanding of the unique physics issueseed implantation. An extensive survey by Preteet al.40 inthe U.S. for prostate brachytherapy revealed alarming vance in the pattern of practice in physics and dosimetry, pticularly in regard to dose calculation, seed assay and timethod of postimplant imaging. Because of the large num

ledesta

apin

lain

a.

elah

64ap

y

a

m

thisived

meicalsedofsi-te

ent

asnt. IIIpy,entinssues

r aini-re

icalely

eis aerhatof

enttiremera-nd

olean-re-onallynt-cti-uessary,eer-s itill:re-ses-

se


of start-up programs at this time, it is essential that the roand responsibilities of the medical physicist be clearlyfined, consistent with the pivotal nature of the clinical phyics component in assuring the ultimate success of prosbrachytherapy.

It was against this background that the Radiation TherCommittee of the American Association of PhysicistsMedicine formed Task Group No. 64, which was charged~1!to review the current techniques in prostate seed impbrachytherapy,~2! to summarize the present knowledgetreatment planning, dose specification and reporting,~3! torecommend practical guidelines for the clinical medicphysicist, and~4! to identify issues for future investigationAlthough high dose rate~HDR! brachytherapy for prostatcancer has certain similarities with permanent seed imptation, the topic is beyond the scope of this Task Group. Treport represents the work of the AAPM Task Group No.The report has been approved by the Radiation TherCommittee and the Science Council.

The dosimetry formalism for interstitial brachytherapwas standardized by the AAPM Task Group No. 43.41 Acode of practice for brachytherapy physics in general woutlined by Task Group No. 56.42 The present report willaddress the clinical medical physics issues unique to per

FIG. 1. Process flow diagram for a preoperatively planned prostateimplant.


s--te

y

nt

l

n-is.y

s

a-

nent prostate seed implant. It should be emphasized thatis a rapidly evolving treatment modality and an area of actinvestigation. Much of our current knowledge in optimizetreatment planning, intraoperative uncertainties, the ticourse of prostate volume change, correlation of radiologstudies of the prostate, and postimplantation analysis is baon research efforts which are still ongoing. The intentionthis report is therefore to guide the practicing medical phycist in successfully implementing or improving the prostaimplant procedure, and to provide a survey of the currstandard of practice in this evolving field.

The remaining parts of this document are organizedfollows: Sec. II provides a practical review of the curretechniques in ultrasound-guided seed implantation; Secreviews the dosimetric aspects of prostate brachytherawith an emphasis on the present knowledge in treatmplanning, dose specification and reporting; Sec. IV contathe summary of recommendations; Sec. V discusses isfor future consideration.

II. REVIEW OF CURRENT TECHNIQUES

A. Overview of contemporary techniques

The major goal of prostate brachytherapy is to delivetumoricidal dose to the cancer-bearing prostate while mmizing urinary and rectal morbidities. The specific aims ato design the optimal treatment plan using 3D anatominformation, to implement the treatment plan intraoperativwith precision, and to analyze the dosimetric outcompostimplantation. Contemporary prostate brachytherapymulti-disciplinary treatment modality, in which each membof the implantation team brings specialized knowledge tpromotes the clinical goal. Figure 1 delineates the flowevents pertinent to the medical physicist in this treatmmodality. The role of the medical physicist spans the enprocess of patient treatment, from the planning volustudy, dosimetric planning, seed preparation, to intraopetive consultation and radiation safety supervision, apostimplant dosimetry.

B. Treatment planning techniques

Computerized treatment planning plays an important rin modern prostate brachytherapy. Careful dosimetric plning leads to smooth and expedient implantation, andduces the likelihood or extent of normal tissue radiatidamage. The process of dosimetric planning is especihelpful to the implant team in the early stages of implemeing the prostate brachytherapy program. It allows the prationers to contemplate the technical and dosimetric isspresented by each case and make adjustment, if necesprior to implantation. Although it is time consuming, thplanning process requires the team to give prior considation to the patient’s anatomy and any technical problemmay present. It allows the team to formulate a plan that w~1! provide coverage of the entire target volume by the pscribed dose while keeping the rectal and urethral dowithin acceptable tolerances,~2! control dose inhomogene

ed

s-

t intiftfe

mmmn

mimntethc

te

trart

sthoe

loh

n

tercsinpilyrei

rcntonebpy,ar-

anis

ra-lon-xialhe

bee

ify

ndxist.nter-d tor

h

lyThiss orin then

ofeed

i-

ndngraing.

onure-be

thra

saitsecmram.ra-the

re-edsn,ay


ity, and ~3! keep the implant as technically simple as posible. The following topics are important in planning the implant:

1. Ultrasound volume study

The ultrasound volume study used to plan the implanusually obtained no earlier than 2–3 weeks before implation, in order to limit changes in the prostate, particularlythe patient is under hormonal therapy. If it is not possiblecomply with this time interval~e.g., due to a shortage oseeds!, a second volume study and computerized treatmplan may be performed before implantation. The volustudy consists of consecutive axial images obtained at 5intervals from the base of prostate to the apex, with the teplate hole pattern superimposed on each image. Practitiowho enlarge the planning target volume~PTV! beyond theprostate often start 5 mm superior to the base and end 5inferior to the apex. In either case, a sagittal ultrasoundage is often obtained for base-apex length measuremeassure that the proper number of slices are obtained. A mber of the physics staff is usually present to ascertain thatpatient is set up in such a manner that a satisfactory planbe developed from the volume study. Specific paramethat are checked include:~1! the angle of elevation of thepatient’s legs in the stirrups;~2! the alignment of the ultra-sound probe with respect to the prostate in all of the ulsound images, such that implant needles, which are inseparallel to the probe, do not traverse the rectal wall;~3! thesuperposition of the template hole pattern on the contourthe prostate. In particular, the most posterior aspects ofprostate need to be within or very close to the posterior rof template holes in order to adequately cover the prostatthe prescribed dose.

The planning volume study ideally includes adequatecalization of the prostatic urethra on each axial slice. Tseed configuration is then designed to avoid implantatioor near the location of the prostatic urethra.

2. Pubic arch

The first consideration in the planning process is to demine the degree of pubic arch interference. The pubic amay ‘‘shadow’’ the anterior and lateral portions of the protate, making it difficult or impossible to implant seedsthese locations. If this restriction exists, the brachytheramay angle the template and ultrasound probe assembachieve better needle access. However, the ability to corthis problem is limited. Severe pubic arch interferenceconsidered a contraindication for performing the implant.

Both CT and TRUS have been used to detect pubic ainterference. In the CT-based technique, the largest extethe prostate is manually projected onto the axial slice ctaining the pubic arch. If significant overlap exists betwethe two structures, pubic arch interference is likely toencountered. A shortcoming of this technique is that thetient is not in the lithotomy position during the CT studthus the relationship between the prostate and the pubicmay be slightly different during implantation. In the TRUS


-

sa-

o

ntem-

ers

m-to

m-e

anrs

-ed

ofe

wby

-eat

r-h

-

sttocts

hof-

nea-

ch

based technique, the ultrasound probe is first moved toaxial slice on which the pubic bone-soft tissue interfacevisible. The location of the pubic arch is traced on the ultsound screen using the cursor. The probe is then movedgitudinally to visualize the entire prostate on successive aslices, with the tracing of the pubic arch overlaid on timages. The advantage of this technique is that it cancombined with the volume study, with the patient in thtreatment position. However, it is harder to precisely identthe pubic arch on ultrasound compared with CT.

3. Seed distribution

Different types of seed distributions are in current use aa consensus on the optimal seed distribution does not eThe classic approach is to space the seeds 1 cm apart, ceto-center, throughout the prostate. This approach, referreas uniform loading, requires a higher number of lowestrength seeds~typically 0.4 to 0.5 U seed for125I, 1.2 to 1.5U seed for103Pd!, and is characterized by relatively higdoses in the center of the prostate. Inmodified peripheralloading, some seeds in the central portion of a uniformloaded implant are deleted to reduce the central dose.may require increasing the strength of the remaining seeddecreasing the needle to needle or seed to seed spacingperiphery.Peripheral loadingis an alternative approach iwhich the seeds are preferentially limited to the peripherythe prostate. This requires a substantial increase in sstrength~typically 0.75 to 1.0 U/seed for125I, 2.0 U/seed orhigher for 103Pd!. The end result is to produce a dose minmum ~albeit above the prescribed minimum dose!, instead ofa dose maximum, at the location of the urethra.

4. Urethra

The prostatic urethra is readily visualized on TRUS aCT studies when a Foley catheter is left indwelling duriimaging. Alternatively, aerated gel injected into the urethcan act as a contrast-enhancing agent under TRUS imagIn order to plan the treatment to avoid direct implantatinear the urethra or to calculate the dose received by thethra, the entire length of the prostatic urethra needs tovisualized. Based on the study by Wallneret al.,38 it appearsthat the maximum urethral dose and the length of the urethat receives greater than 360 Gy~converted from 400 Gy ofpre-TG43 dose for125I! are significantly correlated withRTOG grade 2–3 urinary morbidity. In another study, Deet al.43 reported that acute urinary morbidity in 117 patientreated with125I implants correlated with the dose-volumhistogram of the prostate as well as doses delivered to 52

of the urethra as measured by the dose-surface histogAdditional studies incorporating more patients, differentdionuclides and various seed loading patterns will aiddetermination of the dose tolerance to the urethra.

Treatment plans are commonly devised to limit the uthral dose whenever possible. To accomplish this goal, seare not placed in close proximity to the urethra. In additiosome seeds in an otherwise uniformly loaded implant mhave to be deleted to achieve this goal.

teheanma

tiodti

vetaibri

lyhacrinerth

tumlea

mn

enyeilled

os

thantieo-nsser oorbre

orgle-

in-

firstStolan-

.

eed


5. Rectum

The anterior wall of the rectum is adjacent to the prostawhich makes it difficult to deliver the prescribed dose to tposterior periphery of the prostate without deliveringequivalent dose to the most anterior portion of the rectuPlacing the seeds too close to the rectal wall may increthe risk of ulceration, whereas the extreme posterior porof the prostate may be underdosed if the seeds are placefar away. Particular attention is given to the recto-prostainterface in planning the implant. The physicist aims to cothe entire prostate while keeping the volume of the recwall that receives the prescribed dose as small as possSpecial care is taken where seeds must be positioned nearecto-prostatic interface, especially if peripheral loadingemployed using higher activity seeds.

According to a dosimetric study by Wallneret al.based onCT scans taken 2–4 h after implantation,38 the rectal surfacethat receives greater than 90 Gy~Converted from 100 Gy ofpre-TG43 dose for125I! appears to correlate significantwith rectal bleeding or ulceration. This study suggests teither the dose-surface histogram or the amount of the rewall that receives greater than 90 Gy is a useful parametedosimetric planning and analysis. Again, further studiescorporating more patients, different radionuclides and diffent seed loading patterns will aid the determination ofrectal dose tolerance.

6. Dose margin

Due to seed placement uncertainties that are inherenthe implant procedure, the percentage of the prostate volthat is covered by the prescribed dose is almost alwaysthan planned. Thus, if the prescribed dose and coverageto be achieved it may be necessary to ‘‘over plan’’ the iplant. This is achieved in a variety of ways: by using a planing volume that is larger than the prostate volume~which isalso justified by the known incidence of extracapsular extsion of disease!, by increasing the total activity implanted babout 15%, or by increasing the number of seeds or sstrength until the prescribed isodose line lies several mmeters outside the prostate. All of these methods effectivconstitute a planning integral dose escalation. Thus, thecision to plan a dose margin is tied to the prescribed ditself.

7. Intraoperative planning

Although preoperative dosimetric planning has beencommunity standard in modern prostate brachytherapy,two-step process from the planning volume study to impltation contains several sources of uncertainties. The paposition in the planning volume study is difficult to reprduce in the operating room, leading to ad hoc modificatioof the treatment plan at the time of implantation. Anesthemay result in relaxation of pelvic musculature and conquent change in prostate shape compared to the contoutained in the volume study without anesthesia. Furthermthe prostate may undergo volume change in the intervaltween planning and implantation, which invalidates the p


,

.sentoocrlle.thes

ttalin--e

toe

ssre

--

-

edi-lye-e

ais-nt

sia-b-

e,e--

operative plan. Appreciable change in prostatic volumeposition can occur following the insertion of stabilizinneedles, introducing yet another source of error in impmenting the preoperative plan.

These problems can be alleviated by the technique oftraoperative optimized treatment planning.44 With the patientanesthetized and in the treatment position, the prostate isstabilized using implantation needles. A complete TRUvolume study follows, from which images are transferredthe planning computer and segmented. The treatment p

TABLE I. Equipment requirement for the prostate seed implant program

Mick applicator technique Pre-loaded needle technique

Capital equipmentWell-type ionization chamber Well-type ionization chamberGM or scintillation detector GM or scintillation detectorIon chamber survey meter Ion chamber survey meterComputer treatment planning

systemComputer treatment planning system

Ultrasound unit Ultrasound unitStabilization device/attachment Stabilization device/attachmentFluoroscopy unit Fluoroscopy unitMick applicator

Supplies and consumablesLoading block, cartridges Needle box,~optional! needle

loading deviceSeed carrier Seed sterilization containerMick-compatible needles Needles~Optional! stabilization needles ~Optional! stabilization needlesReverse action tweezers Reverse action tweezersRadioactive seeds Radioactive seeds

Spacers and bone wax

FIG. 2. Process flow diagram for an intraoperatively planned prostate simplant.

trithom

alannt

sor

fieths

ueth

inpiilisofn

bngp

rtd

,bste

ndh

beoom

ding

lav

inin

msth

inonseionhesth

letion

ofand

et-singthe

y,the

sia-eedap-omigh-the, to

gy,

rces

ns

im-

his

ltingies

7.4i-is

p-a

0.4

97

ph-hinard


ning system is invoked to produce optimized dosimeplan, which is subsequently reviewed and approved byphysicist and the radiation oncologist in the operating robefore seed placement takes place.

Intraoperative computerized planning allows several mjor steps to be streamlined. Specifically, preoperative pning can be reduced to a screening of pubic arch interfereand a volumetric measurement to estimate the maximumtal source strength required. Seed inventory can be condated to the same batch for a group of patients, whichduces the variance for waste. Figure 2 shows the modiflow of events pertinent to the medical physicist underintraoperative technique. Most notably, intraoperative dometric planning is a critical step in this treatment techniqwhere the medical physicist occupies an important role inintense decision-making process.

C. Equipment and applicators

Equipment for ultrasound-guided prostate implantscludes the ultrasound machine, the rectal probe, the stepdevice/probe carrier, the perineal template, and the stabing mechanism~see Table I!. The ultrasound machine itypically a portable unit, and contains a seed implant sware package such that a grid pattern can be displayed oscreen. The stepping device allows the rectal probe toattached to the stabilizing mechanism while permittimovement in and out of the patient’s rectum in precise steThe needle template has holes accepting 17 gauge ogauge needles, arranged typically in a 13 by 13 matrix, amm spacing. The template may be designed to mountrectly to the rectal probe in some commercial systemswhich case it moves together with the probe, or it maymounted on the probe carrier, in which case it remainstionary with respect to the perineum as the probe is movIn either case, the holes on the needle template correspothe grid points displayed on the TRUS monitor screen. Tstabilizing mechanism immobilizes the entire rectal procarrier/template system against the operating table or flto prevent unintentional motion of the probe and needle teplate during the implant procedure. The template is placeclose proximity to the perineum to minimize needle splayin the target volume.

Ultrasound equipment is now available that can dispthe sagittal as well as transverse planes of the prostateume. This feature has been found to be helpful in identifythe superior prostate capsule to guide individual needlesertion, in visualizing the movement of the prostate voluas needles are inserted, and in confirming that the seeddeposited correctly at the cephalad-most portion ofprostate.45–47

Equipment required for the prostate volumetric studycludes ‘‘stirrups’’ to support the legs and an examinatitable that allows the mounting of stirrups. The patient isup in an extended lithotomy position on the examinattable, with the thighs at approximately right angles to tbody. This rotation of the pelvic bones allows better accto the prostate, and helps avoid needle obstruction by


ce

--

ceo-li-

e-d

ei-,e

-ngz-

t-thee

s.185i-

inea-d.to

e/r,-

at

yol-g-

earee

-

t

ese

pubic arch. It is important that similar stirrups are availabto achieve the same patient position during the implantaprocedure in the operating room, so that the relationshipthe prostate and the adjacent organs can be accuratelyquickly reproduced. The patient is prepared for the volumric study using Fleets enema, and may be catheterized ua Foley catheter with saline injected into the bladder andFoley bulb in order to identify the urethra. Alternativelaerated gel may be injected as a contrast agent to identifyurethra by ultrasound.

A lubricant ~such as K-Y jelly™! is needed to help intro-duce the rectal probe into the rectum. Topical anesthe~such as Lidocaine jelly™! may be needed for sensitive patients. A small amount of saline or coupling jelly should binjected into the condom over the rectal probe for improvimaging quality and positioning of the posterior prostate csule in the ultrasound image. Too much saline in the condmay however cause distortion of the prostate and the neboring anatomy. The amounts of saline injected intobladder, the Foley bulb, and the condom may be recordedbe reproduced in the implant procedure.

D. Source type, assay and preparation125I and 103Pd sources are comparable in photon ener

capsule dimensions and dose distribution.125I and 103Pd areencapsulated in titanium and delivered as sealed sou~‘‘seeds’’!. They are similar in size~4.5 mm30.8 mm outerdimensions for 125I model 6711 seeds and 4.5 mm30.81 mm for103Pd model 200 seeds!. Both 125I and 103Pddecay via electron capture.125I and 103Pd are currently pro-duced by nuclear reactors and cyclotrons, respectively.

The 125I decay scheme results in the emission of photowith energies of 27.4 keV~1.15 photons/disintegration!, 31.4keV ~0.25/dis! and 35.5 keV~0.067/dis!.48 At the time of thisreport, the seed type most commonly used for prostateplants is model 6711, which contains125I in the form ofsilver iodide deposited on the surface of a silver rod. Tsilver rod also serves as a radiographic marker.125I model6711 seeds therefore also emit fluorescent x-rays resufrom photoelectric interaction in the silver rod, the energof which are 22.1 keV~0.15/dis! and 25.5 keV~0.04/dis!.The average energy for all emissions is approximately 2keV, which results in a half value layer in lead of approxmately 0.025 mm. The self absorption of this assemblyapproximately 37.5%. Therefore, contained activity is aproximately 1.6 times the apparent activity. The air kermstrength used for prostate implants is commonly betweenand 1.0 U~0.3–0.8 mCi! per seed. The halflife of125I is 59.4days; ninety percent of the total dose is delivered in 1days.

103Pd emits characteristic x-rays of 20.1 keV~0.656/dis!and 23.0 keV~0.125/dis!.48 The half value layer in lead is0.008 mm. The active radionuclide is plated onto two graite pellets on either side of a lead radiographic marker witthe titanium capsule. Each end of the seed is cupped inw~i.e., it is concave!. This is a salient feature of the103Pd seedsand therefore can be used to uniquely identify103Pd seeds.

t tednethn

f

heoduman

pans

riliot

dsess

haof

aiodlteo

heThom

teioa-

divbvalen

osa

n

tireenlalegh

d innd ais

theatthealsothet is

thenicisessltra-of

ingcar-theiallyom-in-the

ellned

forcers

eat-ach

ingical

gthen

on.bleingrn,

n oft as

pidtiff,

cmic-

-ialois-

nds-is

ax,ithsing


~Other seeds are convex where the two end welds meemain source body.! The self-absorption of the seed is statto be 54% by the manufacturer, indicating that the contaiactivity of 103Pd in each seed is approximately 2.2 timesstated apparent activity. The air kerma strength commoused for prostate implants is between 1.4 to 2.2 U~1.1 to 1.7mCi! per seed. The half-life of103Pd is 16.97 days; 90% othe total dose is delivered in 56 days.

AAPM Task Group No. 56 recommends that 10% of tseeds be assayed.42 There are several seed assay meththat address the special circumstance in which a large nber of loose seeds are contained in a shipment. Seeds cassayed in bulk, or in cartridges.49 In addition, an autoradio-graph can be taken with a large number of seeds to comthe resulting film density. This film method, in conjunctiowith the well chamber seed assay, assures that the seedof uniform strength. Seeds in suture are delivered stethereby complicating the assay procedure. A calibratcheck may be performed on single nonsterile seeds fromsame batch as a given shipment of seeds in suture andered expressly for this purpose. Alternately, a sterile into the standard dose calibrator can be used to directly aseeds in suture, as described by Feygelmanet al.50 and byButler et al.51 The advantage of the latter techniques is tsterility may be maintained while a sufficient numberseeds are assayed.

Unlike seeds in suture, loose seeds are not sterileneed to be sterilized prior to use. The method of sterilizatdepends on the implantation technique. If preloaded neeare used, the seeds are sterilized prior to loading. After silization, the needles are loaded under sterile conditions,ten in the operating room. If the Mick applicator is used, tseeds can be loaded into cartridges prior to sterilization.sterilized cartridges are then taken to the operating roready for use.

Seeds are commonly sterilized in an autoclave. Flash silization can be used, or longer duration steam sterilizatmay be opted if time and availability allow. Flash steriliztion is done in the autoclave at 270 °F~133 °C! at 30 PSI forat least 3 minutes. The conventional autoclave cycle250 °F~121 °C! at 15 PSI for about 30 minutes. Loose seecan be sterilized in the vial/lead pig in which they are delered, with the cap loosened. Alternatively, the vial canuncapped and the open end plugged with cotton. The gracycle is preferable to the vacuum cycle when loose seedssterilized. The difference between gravity and vacuum cycis the drying method. The vacuum cycle uses a strovacuum to achieve drying. The vacuum may displace loseeds from the container, causing potential radiation hazSeeds may also be sterilized using ethylene oxide gas~coldgas!. Cold gas sterilization takes considerably more time ais required for seeds in suture material.

Seed preparation requirements depend on the implantatechnique adopted, i.e., using the Mick applicator or ploaded needles. Table I shows a list of typical equipmrequired for either technique. The needles used for imptation also differ. The Mick TP200 applicator compatibneedles have a blunt needle sheath and a stylet that is sli


he

dely

s-be

re

aree,nheor-rtay

t

ndnesr-f-

e,

r-n

iss-eityresge

rd.

d

on-t

n-

tly

longer than the sheath, with a trocar point. Needles usethe pre-loaded technique have a sharp beveled point, ablunt stylet that is slightly shorter than the needle. A marktypically present on the end of the needle to indicatedirection of the bevel at the tip. This can be helpful in ththe needle track may be made to deflect slightly towardbeveled direction, if so desired. The needles and styletshave centimeter markings to help visually determinedepth of needle insertion, and the length of the needle thafilled with seeds and spacers. The first 0.5–1 cm ofneedle is usually sand-blasted for increased ultrasoechogenicity. If the image of the needle tip on ultrasoundused to infer the actual needle depth, it is important to assthe precise depth where the needle first appears under usonic imaging; otherwise a systematic positioning error0.5–1 cm can occur.

In the Mick applicator technique,125I or 103Pd seeds areloaded into the Mick-compatible cartridges using the loadblock and a pair of reverse action tweezers. The loadedtridges are then screwed into either the loading block orseed carrier and sterilized. Some seeds are commercavailable in pre-loaded plastic cartridge inserts that are cpatible with the Mick applicator. Use of such pre-loadedserts minimizes radiation exposure to the personnel andtime required for loading the cartridges. The user’s wchamber calibration factors may, in this case, be obtaispecifically for such pre-loaded cartridge inserts.

The pre-loaded needle technique requires longer timepreparation. In this case, the radioactive seeds and spaare loaded into sterilized needles as specified by the trment plan. The loading pattern of seeds vs. spacers for eneedle may be printed on a diagram to facilitate the loadprocess. The needle tip is plugged with a piece of surgbone wax or rectal suppository. The length of the wax~ap-proximately 5 mm! should be accounted for when depositinthe seeds into the prostate. The loaded needles areplaced into a sterilized needle box, ready for implantatiVarious needle loading devices are commercially availathat aim to reduce the amount of time required for loadthe needles, permit visual verification of the loading patteand reduce radiation exposure to personnel.

A Geiger–Muller~GM! counter or scintillation detector isused to survey the seed preparation area after completiothe loading process. A running total of the seeds is kepthey are loaded into cartridges or needles.

In early 1995, Amersham Healthcare introduced RaStrand™ in which the I-125 seeds are enclosed within a sabsorbable suture material that maintains the seeds 1apart center-to-center. The suture material is braided Vryl™ ~polyglactin 910! which is stiffened thermally and sterilized by ethylene oxide. The stiffened Vicryl suture materis hygroscopic and softens and swells when exposed to mture from body fluids. If not handled properly, the stramay swell and jam in the implant needle making it imposible to expel the strand from the needle. Therefore, itimportant that the needle be plugged properly. Bone wwhich is often used to seal the tip of needles loaded wloose seeds and spacers, is too hard to expel without cau

thmhehewa

izeo

padr

mn.gtfoa

sis

–3yhathary,h

biolprs

inadla

nngul

aaeaili

iolaeingri

ntosi-ch-ein

edheetteh-the

, andof

edlane

o-seeoley

nyeedsust-en-

rileeld-theedar-

etscli-the

in-ref-berlaced.e oftheedsle iseade-itedodion.le a

ale.s ton.be

e isombe


a mechanical collapse of the Vicryl spacing betweenseeds. Suppositories, such as Anusol-HC™, are a softerterial that partially melts at 37 °C and allows expulsion of tstrand without collapsing the Vicryl. The internal bore of tneedle must be very smooth. Any roughness of needlewill catch the Vicryl fibers and jam the needle.

The technique for use of Rapid Strand can be summaras follows.52 Place 3 suppositories in a 30 ml glass cupbeaker and flash sterilize in a steam autoclave for 3 min140 °C and 2 atm pressure. With aseptic technique, an em18 gauge implant needle is dipped vertically into the clemolten suppository to a depth at least covering the neebevel and preferably covering the burnished echogenicgion of the needle~approximately 5 mm!. Upon removingthe needle vertically, capillarity and gravity equalize to fora liquid column 7–9 mm long that solidifies in 3–10 miOnce the needles have cooled for 5 min, appropriate lenof Rapid Strand are cut and inserted into the needles,lowed by the stylet. After inserting the needle into the ptient, the moisture resistance of the suppository seal pereven though the needle may be retracted and reinsertederal times in achieving the desired location. During this timthe suppository is warming and melting. Approximately 2min at body temperature are required for the suppositormelt sufficiently around the perimeter of the needle so tthe strand can be easily expelled. Two circumstancesmay lead to jamming of the strands within the needleleaving too little time for the suppository to melt sufficientlor striking the pubic bone, which may dislodge or disrupt tsuppository plug.

E. Implantation procedure

Ultrasound-guided prostate implant procedures canperformed in an operating room or an interventional radogy procedure room. After anesthesia, the patient is set uthe dorsolithotomy position, and draped with sterile coveThe perineum is cleaned with Betadine™ and the scrotumretracted by either sewing it to the drapes or by using a slThe ultrasound probe is attached to the stepping deviceinserted into the rectum. With the exception of the neetemplate, these instruments are usually nonsterile, butcleaned prior to use.

The prostate may be first immobilized against lateral aanterior-posterior motion by use of two or more stabilizineedles, inserted under TRUS visualization through carefchosen template positions.53,54 Special prostate stabilizingneedles with a hook-type mechanism are commercially avable, though standard implant needles for stabilizationalso effective.53 In general, three stabilizing needles arrangin a triangular pattern inside the prostate are quite adequHowever, not all brachytherapy practitioners use immobzation needles.

Under the current preoperatively planned implantattechnique, the positioning of the prostate on the tempgrid under TRUS is carefully checked against the treatmplan both before and after the insertion of the stabilizneedles. Offset in the relative position by exactly one g


ea-

ll

dratty

r,lee-

hsl--stsev-e

totate

e

e-in.isg.ndere

d

ly

il-redte.-

ntent

d

width ~5 mm! is easily corrected by shifting the treatmeplan. Other offset amounts need to be remedied by reptioning the patient. Under the intraoperative planning tenique, prostate stabilization is followed by TRUS volumstudy and optimized dosimetric planning, as describedSec. II B.

For either the Mick applicator technique or the pre-loadneedle technique, a needle loading diagram and/or worksis required in the operating room to identify the templacoordinates for needle insertion. In the Mick applicator tecnique, the loading diagram specifies the spacing betweenseeds in each needle, the number of seeds in the needlethe distance of the first seed-drop position from the basethe prostate~or any other reference plane!, i.e., the offsetfrom the reference plane. The diagram for the pre-loadneedle technique specifies the offset from the reference pfor each needle.

The operating table ideally allows the placement of a mbile fluoroscopy unit to visualize the implanted area. The uof fluoroscopy during implantation helps in visualizing thneedles and seeds as they are inserted in relation to a Fbulb filled with contrast media, and the patient’s boanatomy.55 As a needle is inserted into or pulled out of thprostate, the movement of the previously implanted secan be readily seen on the fluoroscopy monitor, and adjment in the needle insertion depth may be made to compsate for such movement.

If needles are to be loaded in the operating room, a stetable or work area equipped with adequate radiation shiing is set up by a member of the physics staff. Even inMick applicator technique, a set of sterile loading and sehandling equipment is kept available in case jammed ctridges need to be reloaded.

During seed placement, the medical physicist interprthe planning information in each incremental step to thenicians, and records the progress of seed deposition inpatient. Typically, the medical physicist provides verbalstructions on the needle coordinates, the offset from theerence plane, and, if using the Mick applicator, the numof seeds and the seed spacing, as each needle is being p

The technique for seed placement requires some degremanual dexterity. In the pre-loaded needle method,needle is withdrawn against the stylet such that the seremain in the same position in the prostate as the needremoved. Advancing the stylet will deposit the seeds ahof their intended locations, while allowing the stylet to rtract with the needle will cause the seeds to be deposbehind their intended locations. The Mick applicator methrequires the clinician to keep track of each seed depositDesired seed spacing is achieved by retracting the needknown number of steps on the applicator’s preset scWhen the needle is retracted too rapidly, the seed tendfollow it due to the suction created by needle retractioWhen the stylet is advanced too rapidly, the seed caninjected beyond its intended location in the prostate. Cartaken to ensure that only one seed is allowed to drop frthe cartridge at a time, otherwise multiple seeds willplaced at one planned location.

irs

cta

tario

etansisebue

ysgs

la

iaotothrean

acbnth

latinthotoyioicaatheneten

hyraim,ri-

re-theeed

osedpre-

r-nepo-tateaph.hento

n beet-the

sual-

ose

ctn.nt

pro-as

lityops.ay

aly-u-

rid

ones.CTilityhatac-evi-

theased

therec-osef cir-e


It is often found that preoperative treatment plans requminor or major modifications. Typical scenarios that causuch modifications include discrepancies between expeand actual needle placement, detection of the urethra orjacent rectal wall close to the intended seed location, prosvolume/position change, or the need to compensate for pseed misplacement. The medical physicist brings to theerative procedure a special understanding of the dosimimpact of any such modification, and thus plays an importrole in evaluating each circumstance. It should be stresthat while real-time deviation from the preoperative planoften unavoidable for adequate dose coverage, excessivadhoc deviation can lead to severely suboptimal dose distrition. To acquire a quantitative understanding of the dosimric characteristics of a preoperative plan, the medical phcist may wish to simulate a number of such ad hoc chanon the planning computer prior to implantation. This is epecially helpful in the early stages of a prostate seed impprogram.

A running total of the seeds and needles implantedrecorded on the operative worksheet. Any real-time devtion from the planned seed deposition is annotated as itcurs. In the Mick applicator technique, it is good practicereconfirm the running total of seeds deposited every timecartridge is emptied, which serves as a checksum for corseed drop. This is to be facilitated by loading a constnumber of seeds per cartridge~except the last cartridge!.Note that the checksum method does not guard againstdental placement of multiple seeds at the same location,will uncover the problem soon after it occurs. A radiatiosurvey should be made for each used needle to confirmno seed is unintentionally left inside the needle.

After seed placement, the urologist usually performscystoscopy to find and retrieve any loose seeds in the bder. A lead seed container is kept available in the operaroom, for use in the event that seeds are retrieved frombladder, or that seeds are accidentally dropped on the flA GM detector or a scintillation detector is kept availablelocate misplaced seeds, and to conduct radiation survethe implantation area following the procedure. The radiatsurvey includes the floor, waste, linen and all used appltors. All seeds brought to the operating room must becounted for by the implantation worksheet and the seedsremain in the possession of the medical physicist at theof the procedure. A properly calibrated ion chamber survmeter is used to measure the maximum exposure rate asurface and at 1 m from the implanted patient for documtation.

F. Postimplant dosimetry

1. Rationale

The quality of prostate seed implants is, as in all bractherapy, dependent upon the skill and experience of the ptitioner. Because patients differ in their anatomy, someplants are technically more difficult than others. Hencevariation in implant quality may occur, even for an expeenced practitioner.


eeedd-teorp-rict

ed

-t-i-es-nt

is-c-

ectt

ci-ut

at

ad-ge

or.

inn-

c-atd

ythe-

-c--a

Prostate implants are generally planned to deliver a pscribed minimum dose. However, it has been shown thatminimum dose planned can rarely be achieved due to splacement errors which are inherent in the procedure.56–58

Furthermore, postimplant edema can further reduce the ddelivered by the implant.59–61 Hence it cannot be assumethat the patient would receive the dose prescribed in thetreatment dosimetric plan.

Postimplant dosimetric evaluation was traditionally caried out using multiple radiographs. Although such plafilms are adequate for reconstruction of the relative seedsitions, they cannot provide the dose delivered to the prosbecause the prostate cannot be visualized on a radiogrPostimplant dosimetry was limited to a calculation of tmatched peripheral dose~MPD!, a parameter that has beeshown to be an unreliable indicator of the dose deliveredthe prostate.62

The dose delivered to the prostate and other organs cadetermined by performing a postimplant CT-based dosimric analysis. The advantage of CT-based dosimetry is thatprostate and other organs, such as the rectum, can be viized. This capability allows dose-volume histograms~DVHs!to be generated, which provide detailed information on dcoverage and implant quality.

At present, a postimplant CT study is the most diremethod for carrying out quantitative dosimetric evaluatioCT-based dosimetric evaluation is particularly importaduring the early stages of a new prostate seed implantgram to aid the team in progressing up the learning curvequickly as possible. Continuous evaluation of implant quapermits improvement in techniques as the program develOtherwise, problems which compromise implant quality mgo undetected and be perpetuated indefinitely.

2. Technical issues

The necessary steps in performing a CT-based dose ansis are~1! outlining the prostate volume for dosimetric evalation on each CT image,~2! localization of each seed,~3!calculation of the dose to each point in a 3D matrix of gpoints in a selected volume which includes the prostate,~4!generation of isodose curves which can be superposedeach CT image, and~5! generation of a DVH for the prostatas well as dosimetric information for the critical structure

A seed 4.5 mm in length often appears on adjacentimages spaced at 5 mm intervals, therefore a useful facin dosimetric analysis is to permit identification of seeds tappear on multiple adjacent CT images. This is usuallycomplished by superposing the seed location from the prous image onto the image being analyzed.62 Seed redun-dancy algorithms are also helpful, which can reduceseeds to the number actually implanted using distance-bredundancy likelihood analysis.

A complete CT-based dosimetric evaluation includesdose delivered to other organs, such as the urethra andtum. However, there are no standards for specifying the dto these organs, and each case presents a unique set ocumstances. It is very difficult, if not impossible, to defin

eryra

n-ueeomm

rothvth

in

te

sietiv

llt

-s

tarineaw

noly

arlinteter

yte

mtmT

mnt

rs

3i-iate.

thein

helantdiesbeaes

bleos-d

ot-be-tricsor

ob--

T-

icalon

sed

ateg a

n-aremeto

ententt lede-in-

cencyntion


the urethra on a CT image unless there is a Foley cathetthe urethra. Distension of the rectum can cause variabilitassessing the rectal dose due to the typical large dose gent in this region.

The determination of the dose to the prostate frompostimplant CT scan is nontrivial. A major problem is defiing the prostate volume accurately on the CT images. Olining the prostate on CT involves subjective judgment bcause the prostate is not well resolved from other adjacsoft tissue structures. As a result, the volume derived frthe CT scan is generally larger than that of the TRUS volustudy used to plan the implant.63–65 The problem this pre-sents is that the dose coverage will be evaluated for a ptate volume which is larger than that used in planningimplant. As a result, the percentage of the ‘‘prostate’’ coered by the prescribed dose will generally be less thanplanned.

The most notable difficulties in defining the prostateCT images have been described63 as ~1! an inability to dis-tinguish the posterior portion of the prostate from the anrior wall of the rectum on noncontrast CT.~2! a tendency toconfuse the posterior-inferior~apical! portion of the prostatewith the anterior portion of the levator ani muscles, and~3! atendency to include portions of the neurovascular bundlepart of the prostate volume. Because of these difficultdefining the prostate requires a certain amount of subjecjudgment.63

Another problem is postoperative edema, which typicaincreases the prostate volume by 40 to 50% compared topreoperative volume.59,60,66If the postimplant CT scan is obtained immediately after the implant is performed, the domay be underestimated. The edema increases the disbetween the seeds, as well as the volume, thereby lowethe dose rate. On the other hand, if the CT study is obtaiafter the edema has resolved, the dose may be overestimbecause the decrease in dose rate while the prostateedematous is ignored.

The impact of edema on the postimplant dosimetry isyet well defined. However, two factors which intuitivecontribute are:~1! the magnitude of the edema and~2! themargin used in planning the implant. An example of a mgin is an implant planned so that the prescribed isodoseis a few millimeters outside the periphery of the prostaSuch margins are created by the practice of increasingplanned source strength by approximately 15% to compsate for seed placement error.63 One would expect a greatepercentage of the edematous prostate to be covered bprescribed isodose line when such a margin is incorporainto the plan.

The scheduling of postimplant imaging studies is an iportant quality assurance issue because of the effecedema on the postimplant dosimetry. However, the optitime for obtaining the CT scan has not been established.optimal time for imaging125I and 103Pd implants will differbecause their half-lives are different. The duration of edeis a key factor in determining the optimal timing. A recestudy60 based on serial CT scans shows that the edemasolves exponentially with a half-life of from 4 to 25 day


inindi-

a

t--nt

e

s-e-at

-

ass,e

yhe

encengdtedas

t

-e.

hen-

thed

-ofalhe

a

e-

~mean: 9.3 days!. Using the mean edema half-life of 9.days, the edema will typically resolve to 12.5% of its orignal value in 28 days. This would appear to be an approprtime to image an125I implant because of its 60 day half-lifeHowever, the situation is not so clear with103Pd because ofits much shorter 17 day half-life.

Although the methodology has not yet been perfected,TRUS volume study may ultimately become a useful aiddefining the prostate volume in the CT study.65 This is par-ticularly true if the postimplant CT scan is obtained after tedema is resolved so that the preimplant and postimpvolumes can be assumed to be equivalent. If both stuwere imaged at 5 mm intervals, the TRUS study shoulduseful in identifying the apex in the CT study even beforemethodology for registering TRUS and CT images becomavailable. Fusion of CT and MR images may also be a viasolution, as MR provides adequate visualization of the prtate while CT provides localization of the implanteseeds.66,67

These numerous difficulties and technical challenges nwithstanding, the standards for seed implant quality areing defined in terms of quantitative CT-based dosimeevaluation. Willins and Wallner68 reported that, for CT scanobtained on the day of implantation, coverage of 80%more of the target volume by the prescription dose is prably adequate. Biceet al.69,70 have conducted extensive review of postimplant dosimetry using a wide range of Cbased quality assessment parameters. Stocket al.71 foundthat dose was the most significant predictor of biochemfailure in a multivariate analysis using dose, PSA, Gleasscore and stage in 134 patients treated with125I implants. Adose response was observed at a level of 140 Gy in D90, thedose that covers 90% of the target volume under CT-bapostimplant evaluation. Patients receiving a D90 less than140 Gy had a 4-year freedom from biochemical failure rof 68%, compared to a rate of 92% for patients receivinD90 greater or equal to 140 Gy (p50.02).

The clinical correlation of dosimetric evaluators is an ogoing effort. Practitioners of prostate brachytherapyurged to carefully document the methodology and the ticourse for each set of postimplant dosimetry, in orderpreserve its predictive value.

III. REVIEW OF DOSIMETRIC ASPECTS

This section reviews the dosimetric aspects of treatmplanning and postimplant analysis relevant to permanprostate brachytherapy. The historical circumstances thato the adoption of the ‘‘160 Gy prescription dose’’ are dscribed, so that contemporary practitioners can make antelligent judgment with regard to the past clinical experienbased on the Memorial nomograph. Dosimetric consistewith the AAPM Task Group No. 43 formalism is agaistressed. Methods for dosimetric evaluation and optimizaare summarized.

, teo

taat-

bothsuasoalpr

d

reth

teG

owg’m

t od

nu

edtovo

ee

imllytht

thtc

torostono

ecmichv

ichiedsope-a

els,the

ofe

lingtaat,

on-nc-

.2,eenbe

de-

ing

iedllered

y’’-pite

onalen-

ma


A. Historical perspective on dosimetry

Although 222Rn seeds,198Au seeds and even192Ir seedshave been used in permanent implants of the prostatehistorical background of clinical and physics techniques ptinent to the present report really began with the initial use125I seeds for this purpose at New York’s Memorial Hospiin the late 1960’s.72 These implants were performed usingretropubic approach, following a midline incision and bilaeral lymphadenectomy. Ideally, needles were inserted a1 cm apart and parallel to one another, avoiding the ureand stopping short of the rectum by sensing needle preson a finger in the rectum. Each needle was withdrawn at le0.5 cm before the first seed was inserted. The dimensionthe prostate were assessed in the plane perpendicular tneedles as the distance between peripheral needles andthe needle direction by subtracting the average needletrusion from the overall needle length~15 cm!. The totalapparent activity~in mCi! to be implanted was determineby multiplying the average dimension~in cm! by an empiri-cally derived factor of 5. Implementation of this proceduwas eventually facilitated by a nomograph that specifiednumber of seeds~of known strength! to be implanted andtheir approximate spacing within the target.3

The dose associated with an implanted activity demined in the above manner was believed to be about 160and was considered to be the minimum effective dose. Hever, an early evaluation of the ‘‘dimension averaginmethod had shown that, on the basis of Quimby voluimplant data,73 for which the cumulated activity~mg h! perunit dose is approximately proportional to the square roothe volume treated, dimension averaging may be expecteproduce a peripheral dose roughly proportional to the mione-sixth power of volume.5 Alternately, if Manchester vol-ume implant data74 had been invoked, where the cumulatactivity per unit dose is explicitly stated to be proportionalthe two-thirds power of volume, the expectation would hainvolved a dose proportional to the minus one-third powervolume. In either of these conjectures, the premise is that125Iseeds display a dose-rate fall-off with distance from the ssimilar to that from a radium or radon source.~Dimensionaveraging had, in fact, been used earlier for radon seedplants.! Although we now know that assumption to be totaunjustified, it is nevertheless clear that adherence tooriginal dimension-averaging method of planning leadssmaller doses for larger target volumes. During the timeplanning method was in use, it was not really appropriatesuggest that it resulted in delivery of a given dose, sintarget volumes~e.g., prostates! varied significantly in size.

Before the advent of CT imaging, there was no wayevaluate the minimum peripheral dose received by the ptate, since the prostate capsule was not seen on the poplant stereo-shift radiographs from which dose calculatiwere generally performed. In order to provide some formfeedback to the brachytherapy clinicians that would reflthe extent to which implant goals were achieved, it becacustomary at Memorial Hospital to report the dose for whthe isodose contour volume was the same as the target


her-f

l

utrarestoftheongo-

e

r-y-

’e

ftos

ef

d

-

eoisoe

s-im-sfte

ol-

ume inferred from dimensions measured at surgery.75 Thetarget was usually approximated as an ellipsoid, for whthe volume is the product of the three dimensions multiplby p/6. Because it involved matching volumes, the dosereported subsequently came to be called the ‘‘matchedripheral dose’’ or MPD. It was obtained by interpolation intable of volumes computed at uniformly spaced dose leveach volume representing the sum of voxels for whichdose was greater than the specified dose.

The MPD concept proved helpful in later modificationsthe original ‘‘planning’’ nomograph to take into account thdifference in dose rate falloff with distance between125Iseeds and222Rn seeds.6,76 MPD values evaluated for actuaimplants were found to decrease significantly with increastarget volume.77 From the slope of the line fitted to such daon a log–log plot, it was possible to derive the exponent thif applied to the average dimension in a modified dimensiaveraging method, would result in a constant dose as a fution of target volume. This exponent was found to be 2where the increase over the 2.0 value that would have bimplied by Manchester volume implant data was taken todue to the much lower penetration of125I photons relative tothose of222Rn. The same reasoning was later applied tovelop a similar nomograph for103Pd seed implants.4 For103Pd, with photons of even lower energy, the correspondexponent of average dimension was 2.56. For both125I and103Pd nomographs, the ‘‘constant-dose’’ formula was applonly for average dimensions greater than 3 cm. For smaimplants, the original dimension-averaging rule was allowto stand and the dose increased for decreasing volume.

With respect to the ephemeral but pervasive ‘‘160 Gprescription dose for125I permanent implants, a further observation of interest is that it seems to have survived in sof major changes in125I dosimetry. At the time it was firstproposed, dose calculations were using a one-dimensilookup table of dose rate times distance-squared with an

TABLE II. Average dose at total decay from a source with an air kerstrength of 1 U using the point source approximation.

Distance~cm!

Dose at total decay~Gy!

125Imodel 67111985 NISTStandard

125Imodel 67111999 NISTStandard

103Pdmodel 200TheraSeed®

0.5 70.02 77.98 20.191.0 16.83 18.74 3.911.5 6.93 7.71 1.332.0 3.50 3.90 0.562.5 1.97 2.19 0.273.0 1.18 1.32 0.133.5 0.74 0.83 0.074.0 0.49 0.54 0.044.5 0.33 0.37 0.025.0 0.23 0.26 0.015.5 0.17 0.196.0 0.12 0.146.5 0.09 0.107.0 0.07 0.08

D

rgan

v

il

tsa

-

g

elt

den

emeot

etaa

tho

acm

ctnge

g to

antsng43by

so

ofli-gthrder.

%iongyndeon-In

0.98ndsed

ort-an-in

n-el-thein

cifi-nce

ase% of,by

er-r20lini-ned

hePDhe

b-ticundoor


try at a distance of 1 cm of 1.7 cGy cm2 mCi21 h21. Thisvalue had been arrived at indirectly, making use of TLmeasurements of relative dose vs distance in Mix-D phantommaterial. It was considered to be the quotient of total eneemission per mCi h and the product of phantom densitythe volume integral of the relative measurements.78 It waschanged, in 1978, to 1.10 cGy cm2 mCi21 h21 on the basis ofsubsequent measurements and calculations79 that includedaveraging the anisotropy~albeit in air! over 4p solid angle.MPD values were reduced to 65% of what they would habeen using the previous table.

B. Dosimetry data revision

The history of125I dosimetry has been reviewed in detaby both the AAPM Task Group 43~TG43! report41 and thereport of thead hoc Committee of the AAPM RadiationTherapy Committee on125I sealed source dosimetry.80 Asindicated by Kuboet al.,80 two separate but related evenneed to be considered in the discussion of dosimetry drevision for125I seeds:

~1! The adoption of the AAPM Task Group No. 43 recommended dosimetry data for model 6711125I seeds, whichdiffers significantly from the dosimetry data of Linet al.81

~2! The revision of the NIST calibration standard for thtitanium-encapsulated125I seeds, including both mode6702 and model 6711 seeds. This revision causesreported air kerma strength value for a calibrated seebe approximately 10% lower than that under the currNIST calibration standard.

When applied to the dosimetry of permanent prostate simplants, the isotropic point source approximation is comonly used for the dose calculation model. At distancrfrom the center of the source, the dose delivered at tdecay from an125I or 103Pd seed is

D5D031.4433T1/2, ~1!

where the initial dose rateD0 is given by the TG43 formal-ism

D05SkLg~r !fan

r 2 , ~2!

whereSk , L, g(r ), andfan are the air kerma strength, dosrate constant, radial dose function, and anisotropy consrespectively. Table II gives the dose at total decay from125I or 103Pd seed with an air kerma strength of 1 U, usingpoint source approximation. Comparison with the old expsure rate formalism shows that, for a model 6711125I seed ofgiven strength, the dose calculated using the TG43 formism and dosimetry data is approximately 11% lower at 1from the center of the seed in water. Luseet al.82 furthercompared the isodose distribution for the implant of a 35prostate, using 88 seeds and 20 needles. They found thacontiguous volume of 53 cc receiving 160 Gy, using the Liet al.dosimetry data, would receive 144 Gy when the dos


yd

e

ta

hetot

ed-

al

nt,ne-

l-

cthe

is

calculated using the TG43 dosimetry data, correspondinthe 11% difference observed earlier. Luseet al. thereforerecommends that the prescription dose in prostate implusing125I seeds be modified downward by 11%, prescribi144 Gy instead of 160 Gy, when the dosimetry data of TGis used in the dose calculations. Another empirical studyBice et al.83 based on similar comparative analysis alreached the same conclusion.

While there is no timeline constraint on the adoptionthe TG43 dosimetry data, early adoption of TG43 will facitate smooth implementation of the revised air kerma strenstandard by NIST. The revision of the NIST I-125 standawill need to be carried out in concert with the manufacturThis requires a change of the dose rate constantL in theuser’s treatment planning system. Loevinger84 has reportedearlier that this revision would lead to approximately a 10increase of the dose rate constant. The AAPM RadiatTherapy Committee Ad Hoc Subcommittee on Low-EnerSeed Dosimetry reviewed the available data arecommends85 that, for I-125 sources calibrated using th1999 NIST air-kerma-strength standard, the dose rate cstants should be revised upward by a factor of 1.114.particular, it is recommended that dose rate constants ofcGy/h-U and 1.04 cGy/h-U be used for model 6711 amodel 6702 seeds calibrated using the NIST 1999 reviair-kerma-strength standard.

C. Dose specification and reporting

Consistency in dose specification, prescription and reping is an important step towards establishing a uniform stdard of practice in prostatic brachytherapy. Early effortsthis area86–90 were exclusively limited to idealized represetations of the target using cubic, cylindrical, spherical orlipsoidal volumes. However, these investigations markeddeparture from built-in target-size dependencenomograph-based planning and prescription toward specation or prescription of a desired dose. The early experiewith CT-based planning and evaluation for125I prostate im-plants at the Memorial Sloan Kettering Cancer Center wreported by Royet al.62 In their study, the peripheral doswas defined as the isodose surface that encompassed 99the target volume, or D99. By analyzing 10 implant casesthey showed that the actual coverage of the target volumethe peripheral dose ranged from 78% to 96%, with an avage coverage of 89%. More recently, Willins and Wallne68

published a follow-up study presenting the analysis ofunselected implant cases performed by an experienced ccal team. In this study, dose was prescribed to the planminimum peripheral dose~mPD!; postimplant dosimetry wascarried out using CT taken on the day of implantation. Tactual coverage of the target volume by the prescribed mranged from 73% to 92%, with an average of 83%. Tactual D100 delivered to the target volume was 43%68%~61 SD! of the prescribed mPD. The authors identified sujectivity in interpreting postimplant CT scans and prostaswelling as extraneous uncertainties, which would compowith any seed placement errors to result in apparently p

os

enaaoa

edi

lape

ethh

alsThtiothi

chthowPimatl

e-sa

ini

ecn

onuat

onde

rrin

pedon-

lydis-the

m-

andreon.

dernalr inr-etryiffi-nts.on-

.

he-umn

non-oflant

es

ari-uld

zedosethe

bleof

he-


coverage by the mPD. The reliance on the mPD for dprescription and reporting was discussed by Yuet al.56

Based on simulated distributions of common seed placemerrors, they concluded that generally 90% of the origiPTV could be covered by the prescribed mPD when theerage dimension of the PTV was greater than 3 cm. No csistent pattern was found for the magnitude of underdosbetween the planned mPD and the realized D100; however,underdosage from the planned mPD to the realized D100

could easily exceed 20% due to common seed displacemAlthough the mPD as a dose specification parameter

plays excessive sensitivity, it is the most direct measuredosimetric coverage under 3D image-based treatment pning. Less sensitive dosimetric parameters have beenposed for prescription and/or reporting, including the nminimum dose,89 the average peripheral dose,88 the harmonicmean dose,90 and the widely cited matched peripheral dosHowever, these parameters by definition do not provideessential information on isodose coverage of the PTV. Tpractice of prescribing the treatment dose to the mPD isconsistent with modern external beam radiation therapy.problem of achieving consistency between dose specificafor prescription and dose reporting is closely related touncertainties in defining the PTV, which are discussedmore detail in Sec. III D.

It must be concluded at this time that the present teniques for permanent seed implantation do not yet allowplanned mPD to be reproduced with any consistency; hever, the percent coverage of the PTV by the planned misodose surface is a reasonably consistent indicator ofplant quality. Furthermore, the mPD required for cliniccontrol of disease and avoidance of morbidity is currenunknown.

With regard to regulatory compliance, this body of rsearch work strongly suggests that apparent underdofrom the prescribed mPD to the realized D100 for a givenimplant cannot be taken alone as evidence of poor admtration of brachytherapy.

D. Dosimetric uncertainties

Three major sources of uncertainties are now widely rognized: seed displacement, prostate edema postimplation, and difficulty in defining the target volume basedCT. These uncertainties can potentially compound to cagross dosimetric variability and ultimately to affect the trement outcome.

Seed displacement refers to the deviation in the positiof the implanted seeds from the planned locations. Thesimetric impact of seed displacement has been wdocumented.56,57,68,91In particular, Robersonet al.57 classi-fied seed displacement in terms of needle placement esource-to-source spacing variability, and seed splayThese errors arise because~1! the patient position during theplanning volume study is not always reproducible in the oerating room;~2! the prostate volume may have changsince the planning study, particularly for patients under hmonal therapy;~3! prostate movement occurs during impla


e

ntl

v-n-ge

nt.s-ofn-

ro-t

.eeoen

en

-e-D-

ly

ge

s-

-ta-

se-

so-ll

or,g.

-

r-

tation, even with stabilizing needles in place. Use of rigidspaced source strands will partially alleviate the seedplacement uncertainties. However, a complete solution toabove problems is unlikely to result until intraoperative coputerized real-time dosimetry is widely available.

Dosimetric uncertainties as a result of prostate edemadifficulty in defining the target volume based on CT aunique to dose analysis and reporting postimplantatiThese issues are discussed in detail in Sec. II F.

E. Treatment plan evaluation

Methods for dosimetric evaluation are necessary in orto select competing treatment plans during the traditioplanning process, in ranking computer-optimized plans, opostimplant dosimetry analysis. At present, little clinical corelation has been published between the planning dosimevaluators and treatment outcome, due in part to the dculty of achieving the planned parameters in actual implaTo that extent, the dosimetric evaluators are pragmatic cstructs to quantify the treatment plan evaluation process

1. SmPD

The total source strength required to deliver 1 Gy of tmPD,SmPD, is implicity or explicitly used in treatment planning to select the seed distribution that yields the maximtumor dose. Intuitively, the dose distribution within a givePTV is also most uniform whenSmPD is minimized, for oth-erwise some source strength can be removed from theuniform region without reducing the mPD. The conceptSmPD can be traced back to the Manchester system of impdosimetry. For prostate seed implants using125I and 103Pdseeds, the following fitted results provide minimized valuof SmPD as functions of the average dimensiond of thePTV:56,92

125I S50.014d2.05U/Gy-mPD, ~3!

103Pd S50.056d2.22U/Gy-mPD. ~4!

These relationships do not take into account any irregulties in the shape of the target volume, and therefore shoonly be used as idealized estimates ofSmPD. Most clinicaltreatment plans are likely to yield higher values ofSmPD. Forthe purpose of benchmarking comparison with the idealimodel, the three largest orthogonal dimensions of the isodsurface selected for prescription may be used to obtainaverage dimension.

For reference, the following fitted results are obtainafrom Ref. 4 for the total source strength implanted per GyMPD:

125I S50.012d2.2 U/Gy-MPD

~converted to TG43 dose!. ~5!

103Pd S50.36d2.56 U/Gy-MPD. ~6!

For average dimensionsd between 3 and 5 cm, Eqs.~3!–~4!and Eqs.~5!–~6! agree to within 10% for125I or 103Pd. It iswell known that the mPD is substantially lower than tMPD. However, Eqs.~5!–~6! were derived from the dosim

op

tivtte

eitt

ce

3.d-asdim

osnu-c

su

helsint

ffito

isreneonmnnt

ths

tioa-

stoera

thonl

fece.om-t of

p-betionInedex-

era-ningis-

duemess ine-

rtedHoge-

fre-byCIderbe

3Dme-

, inast-

dg-the

d aheles

erethe

een

o


etric analysis of past clinical cases, i.e., arising from posterative analysis, whereas Eqs.~3!–~4! were results of opti-mized ideal seed placement, i.e., arising from preoperaplanning. The apparent agreement of the two sets of fiequations is therefore rather incidental.

2. Dose uniformity

The full-width at half-maximum~FWHM! of the differ-ential dose-volume histogram or the ‘‘natural’’ dose-volumhistogram93 has been used as an indicator of dose uniformfor prostate implants.62,94Greater dose uniformity throughouthe target volume would be associated with a reduFWHM. Roy et al.62 reported 307 Gy673 Gy ~61 SD! inFWHM of the DVH postimplantation for 10 cases in 199Although they did not report the FWHM in the corresponing preoperative plans, the parameter is likely to increfrom the plan to the postimplant dosimetry because seedplacement tends to spread out the peak of the DVH. A spler construct is the uniformity number~UN!, defined as theratio of the mean peripheral dose to the mean tumor dboth calculated as harmonic means in the PTV to avoidmerical instability.90 The UN is about 0.7 for idealized implants, and should be relatively insensitive to seed displament. Dose profiles have also been used to meauniformity in two dimensions through the target volume.95

The notion of achieving dose uniformity is related to tconcept of sterilizing a uniform distribution of tumor celthroughout the PTV, which is also an implicit assumptionmost external beam treatment of prostate cancers. Withpossible exception of minor dose heterogeneity,96 high doseswithin the PTV in excess of that required to produce sucient cell kill is assumed to add risk without benefittherapy.

3. Dose conformity

Dose conformity measures the closeness between thedose surface chosen for prescription and the PTV in thdimensions. It is different from dose uniformity, as demostrated in the following example. If the PTV is a sphere, tha point source located in the center will achieve perfect cformity, but the dose distribution is severely nonuniforthroughout the target. Thus dose conformity alone doesfully define the objectives for optimizing prostate implatreatment plans.

The most common measure of dose conformity isroot-mean-square deviation of the peripheral dose from alected dose level. This evaluator is often used in conjuncwith computerized optimization. Another conformity indictor is the peripheral uniformity number~PUN!, defined as theratio of the mPD to the mean peripheral dose calculatedthe harmonic formalism.90 For planned seed configurationoptimized for125I and 103Pd, the PUN is on average equal0.67. A higher PUN is indicative of better conformity in thtreatment plan. The PUN is likely to undergo severe degdation after seed displacement due to the volatility inmPD. A third parameter proposed in the literature is the cformation number~CN!, which is applicable to both externa


-

ed

y

d

eis--

e,-

e-re

he

-

o-e

-n-

ot

ee-n

in

-e-

beam and brachytherapy.97 When calculated at the level othe mPD, the CN is simply the ratio of the volume of thPTV to the volume enclosed by the mPD isodose surfaFor prostate seed implants, the CN is on average 0.72, cpared to 0.65 for a 3-field external beam boost treatmenthe prostate.

The notion of dose conformity is based on the assumtions that a well-defined and clinically relevant PTV canprecisely identified, and that there is reasonable expectaof delivering the conformal dose distribution as planned.some institutional protocols, the PTV includes an expanddosimetric margin around the true prostate to encompasstracapsular extension and in anticipation of dose degention subsequent to seed misplacement. Treatment planthen aims to conform to the expanded PTV. However, dtortion of the planned isodose surface invariably occursto seed placement uncertainties. Until a technique becoavailable that substantially accounts for these uncertaintiereal time, dose conformity remains an evaluator only of idalized treatment plans.

4. Dose-volume histogram

Compared to single scalar evaluators, the DVH~including‘‘natural’’ DVH ! provides substantially more information foquantitative evaluation of the dose distribution associawith a given plan or actual implant. Variations of the DVconcept include the coverage, external-volume and heterneity indices,21,56,98 and the dose nonuniformity ratio.98 Inparticular, the coverage index~CI! shows the percentage othe target volume covered by any isodose level. In the poperative treatment plan, CI at the mPD dose level isdefinition 100%; in the actual implant, the correspondingat the same dose level should be approximately 90% unthe current implantation techniques, and is expected tohigher with better techniques.56

F. Treatment plan optimization

It was recognized from the early days of image-basedplanning that template-guided prostate implants were anable to computerized optimization. Royet al. first reportedthe Memorial experience of CT-based optimized planningwhich the seed loading patterns were determined by a lesquare method to maximize the dose conformity;99 theneedle patterns were selected on the basis of clinical jument. This method is therefore semi-automatic, in thattreatment planner needs to design a needle pattern~and intheir initial approach, needle orientation! as a starting pointfor computer optimization. Even so, the authors reportefactor of 10 reduction in the planning time compared to tmanual planning experience. The optimal seed loading ruwere explored by Narayanaet al.100 When the effects of po-tential seed displacement and prostatic volume change wtaken into account, peripheral loading appeared to beoptimal strategy.

A number of robust optimization schemes have since bapplied to prostate implants. Pouliotet al.101 used simulatedannealing~SA! to optimize a cost function that took int

genge

e

vetoinA

bleves-io

ee

uth

asicouacgls

akosen

isi-ioprursgr

o

esfe

e

a

ara

cts

rre-m-md-

go-rt of

n,ea-

o-

val-

in-thecalr-s inthe

. In-nd

sedell-snse,

nc-ec-re

sallytiveera-

dardfor-.y-


account both dose conformity and dose uniformity. Thenetic algorithm~GA! has been adapted for inverse plannito minimize SmPD and maximize dose conformity, whilkeeping the number of needles at a customary range.92,102

Chenet al.103 devised an ad hoc method in which one sewas placed at a time untilSmPD was minimized. Overall,these optimization techniques were able to produce improtreatment plans, as measured by the respective evaluawithout extensive human intervention and within 1–15 mrun time on modern computers. In addition, both SA and Gare stochastic, ‘‘intelligent’’ optimization schemes, capaof search for optimality as defined by realistic objectifunctions.104,105 The potential for incorporating seed diplacement uncertainties in such an intelligent optimizatscheme was explored.102,106 It is quite likely that some ofthese techniques will be translated into mainstream treatmplanning and intraoperative dosimetric guidance in the nfuture.

IV. CURRENT RECOMMENDATIONS

Based on the rationale and the empirical evidence olined in the foregoing sections, the AAPM recommendsfollowing practical guidelines as a basis for promotinglevel of quality standards in prostate brachytherapy phythat is necessary to ensure that the anticipated clinicalcomes are reproducible and uniform on a large scale. Prtioners of existing prostate seed implant programs are urto compare these guidelines with their institutional protococarefully evaluate and justify any departures, and mmodifications to their programs if necessary. For new prtate seed implant programs, it is recommended that thguidelines be implemented as part of the quality assuraprotocols.

It is recognized that modern prostate brachytherapymulti-disciplinary effort that involves radiation oncology, dagnostic radiology and urology. Successful implementatand continued improvement of a prostate brachytherapygram rely on effective teamwork and ongoing quality assance review of the entire program. The physics aspectquality assurance as outlined in this section are an intepart of this multi-disciplinary effort.

This section of the document uses three distinct levelsimperatives with strictly defined meanings:

~1! Shall or Must indicates a recommendation that is necsary to ensure a minimum standard of safety and eftiveness in prostate seed implantation;

~2! Shouldindicates a recommendation that is necessarymeet the baseline standard of practice in prostate simplantation;

~3! Recommendindicates an advisory recommendation this to be applied when practicable.

A. Equipment

The medical physicistshouldbe directly involved in theselection, acceptance-testing and quality assurance ofequipment acquired for the prostate seed implant prog


-

d

drs,

n

ntar

t-e

st-ti-ed,e-sece

a

no--ofal

f

-c-

toed

t

nym

~see Table I!. The quality assurance of equipment that affethe dosimetric consequences of seed implantationmust beperformed by the medical physicist.

Imaging: Verification shall be made~e.g., in phantom! toensure that the grid pattern on the ultrasound image cosponds to the physical locations given by the perineal teplate. The fluoroscopy unit used in the operating rooshoulddisplay minimal distortion in a screen area that aequately encompasses the implant region. It isrecommendedto identify and follow a set of acceptance testing and oning quality assurance procedures described in the RepoAAPM Ultrasound Task Group No. 1107 that are relevant toTRUS imaging, especially with regard to spatial resolutiograyscale contrast, geometric accuracy, and distance msurement.

Accessories: Proper functioning of applicators, accessries and stabilizing devicesshould be verified before eachimplantation procedure.

Treatment planning system: The medical physicistshallverify that the treatment planning system reproduces theues shown in Table II for the dose at total decay from an125Imodel 6711 or103Pd seed, calculated using the TG43 datathe point source approximation.40 This test serves as a necessary indication that the planning system complies withdosimetric formalism recommended by TG43. The mediphysicistshall verify that the treatment planning system peforms the correct dose summation at one or more locationa simple configuration of multiple seeds. We endorserecommendations of the AAPM Task Group No. 40108 re-garding quality assurance of treatment planning systemsparticular, the above testsshallbe performed before the computer treatment planning system is put into clinical use, aat each subsequent software release.

Dosimeters: The medical physicistshall establish the cali-bration of dosimeters for the assay of each type of seed uin the prostate brachytherapy program. The user’s wchambershall be calibrated at an ADCL with direct traceability to NIST. Alternatively, individually calibrated seedshall be obtained from the ADCL to establish a calibratiofactor for the particular geometry being used. In either cathe constancy of the user’s dosimetershall be confirmed us-ing a long-lived radionuclide before each use. Proper futioning of the ion chamber survey meter and radiation dettor shall be verified using a long-lived test source befoeach use in the operating room.

B. New radionuclide designs

New supplies of125I, 103Pd or other low energy sourcefor permanent implantation need to become commerciavailable to meet the increasing demand for radioacseeds. However, it must be stressed that for any new thpeutic radionuclide, the dosimetry for the source designmustbe established. Ideally, a national air kerma strength stanshouldbe established, and the parameters for the TG43malismshouldbe measured and independently confirmed

The dosimetry of low energy photon-emitting brachtherapy sources such as125I and 103Pd is sensitive to the

ee

eenacewusa-

tati

edev

Mef

an--

M-

edG

IIIhereolu

xan

oco

-os

o

ys-dlan.ns

atic

ed

is

tion

um-

it isnft

, a

enand

yac-

ues

therti-e ofNo.eal

ra-endss totec-to

eds.

thethernal

ysislti-


source geometry, encapsulation and internal structure duself-absorption effects. These factors can be particularly ssitive to the quality of the manufacturing process during sfabrication. It is inappropriate to use the dose rate constaradial dose functions, anisotropy functions, anisotropy ftors or constants published in the TG43 report for nsource designs. Regarding a source intended for widethe vendorshall have the responsibility to provide a calibrtion of source strength that is traceable to a standard, andmedical physicistshall have the responsibility to ensure ththe clinical dosimetry parameters have been validated bydependent investigators other than the vendor.

C. Seed assay

Radioactive seeds may be obtainable in loose seready-loaded cartridges, or absorbable suture. In whatform the seeds are procured, the manufacturer’s assaymustbe independently confirmed. As recommended by AAPTask Group No. 56,41 a random sample of at least 10% of thseeds in the shipmentshouldbe checked. Discrepancies o3% or more between the mean of the assay and the mfacturer’s calibrationshouldbe investigated. Unresolved discrepancies of 5% or moreshouldbe reported to the manufacturer.

As discussed by the Ad Hoc Committee of the AAPRadiation Therapy Committee on125I sealed source dosimetry, the revision of the NIST standard for125I mustbe takeninto account as soon as it becomes available.

D. Changes to the dose value due to TG43

To promote uniformity in the clinical adoption of thTG43 formalism, it isrecommendedto scale the prescribedose such that a pre-TG43 value of 160 Gy becomes 145This recommendation is based on the discussion in Sec.but with the dose rounded from 144 Gy to 145 Gy. Tclinically optimal dose and the method of prescription anot yet definitive. In cases where a pre-TG43 prescribed dother than 160 Gy needs to be converted to the TG43 vait is recommendedto use the scaling ratio of 0.9.

E. Dosimetric planning

Treatment planningmust be carried out for all patientsprior to the insertion of radioactive seeds. In this contetreatment planning refers equivalently to intraoperative plning or conventional preoperative planning. It isrecom-mendedto generate the isodose distributions superposedthe contours of the prostate in selected planes, and tostruct the DVH for the prostate. It isrecommendedto gener-ate the DVH or ideally the dose-surface histogram~DSH! forthe rectum. It isrecommendedto adequately identify the entire length of the prostatic urethra, and to calculate the dprofile along the urethra.

Prior to implantation, the dosimetric planshould bechecked using an independent procedure or by a secmember of the physics staff, andmustbe reviewed by theradiation oncologist.


ton-dts,-

e,

he

n-

s,er

u-

y.B

see,

t,-

nn-

e

nd

F. Implantation procedure

A member of the physics staffshall be present in theoperating room during prostate seed implantation. The phics personnelmustbe familiar with the treatment plan anthe dosimetric consequences of any deviation from the pIf the implantation technique relies upon preoperative plabased on prior volume studies, the position of the prostgland relative to the template coordinatesmustbe verified inmore than one imaging plane. If deviation from the plannposition is detected, the physics personnelshould evaluatewhether modification to the setup and/or treatment planrequired, and recommend corrective action.

An account of the needles and seeds implantedshall bekept as the procedure progresses. At the end of implantaand after cystoscopy, the physics personnelshall confirm thetotal number of seeds implanted in the patient and the nber of seeds remaining, whichmustadd up to the total num-ber brought into the operating room.

A scintillation detector or GM countermustbe availablein the operating room. For implants using loose seeds,recommendedto survey each needle after it is withdrawfrom the patient, to verify that no seed is unintentionally lein the needle core. At the completion of the procedurecomplete radiation surveymust be conducted, which in-cludes the vicinity of the implant area, the floor, waste, linand all used applicators. The exposure rate at the surfaceat 1 m from the patientshouldbe measured by a properlcalibrated ion chamber survey meter, and documented incordance with pertinent federal and state regulations.

The physics personnelshouldbe familiar with any insti-tutional policies and procedures regarding sterile techniqand the operative environment.

G. Patient release

The medical or health physicistshall routinely review thepatient survey results postimplantation to confirm thatprostate seed implant program continually satisfies all penent federal and state regulations regarding the releaspatients with radioactive sources. NCRP Commentary11, ‘‘Dose Limits for Individuals Who Receive Exposurfrom Radionuclide Therapy Patients,’’ provides additioninformation that may be of use in this context.

For obvious reasons, the institution’s accountability ofdioactive sources for a permanent prostate seed implantat the time of patient release. However, basic instructionthe patient on identifying the seeds and on radiation protion principles should be provided. It is not necessaryrequire the patient to strain urine and return dislodged se

H. Postimplant analysis

A quantitative dose analysismustbe carried out for eachpatient postimplantation. This statement is based onpremise that it is as important to know and documentdose delivered by a permanent seed implant as by an extebeam treatment. The importance of a postimplant analcannot be overemphasized for the purposes of mu

inmy.llu

sm

o

alfoion-ohs

e

m

osb

efoicc

ric

nise

denth

en

erys

tly

l-nthio

telyes

tiveser

dform-

alM

for

in

andoespy.en-eir

n

ysoori-

nedn isd todicthentitybestde-

al-lt,

byon-intfor-

if-thexi-ardareeir

re-


institutional comparison, improving techniques, evaluatoutcome, and identifying patients who might benefit frosupplemental therapy or be at risk for long-term morbidit

The postimplant analysisshouldinclude two-dimensionadose distributions on which the target volume for dose evaation is outlined. In addition, it isrecommendedto constructthe DVH for this target volume, and to document the dolevels that cover 100%, 90% and 80% of the target volufor postimplant evaluation, i.e., D100, D90 and D80, and thefractional volume receiving 200%, 100%, 90% and 80%the prescribed dose, i.e., V200, V100, V90 and V80. Currentliterature suggests that imaging studies for dosimetric evation are ideally obtained 2–3 weeks postimplantation103Pd implants and approximately 4 weeks postimplantatfor 125I implants. However, it is recognized that logistic cosiderations sometimes preclude such uniform timingpostimplant imaging for all patients. In addition, future tecnology may permit immediate postimplant dosimetry assement in the operating room. In any case, the time courspostimplant dosimetric evaluationshouldbe recorded for allpatients.

For dosimetric evaluation performed at the optimum iaging time, it isrecommendedto use D90, in comparison tothe prescribed dose, as an indicator of implant quality in dcoverage. An implant with good coverage is characterizedD90 equal to or greater than the prescribed dose.

It is recognized that such dosimetric analysis is sensitivdependent upon the definition of the target volumepostimplant evaluation. Therefore a consistent radiologinterpretation of the target volume should be used and domented to facilitate future interpretation of the dosimetoutcome.

It is recommendedto construct the DVH or ideally theDSH for the rectal wall. Furthermore, it isrecommendedtoadequately identify the location of the prostatic urethra, ato document the dose to the urethra. We recognize that valization of the urethra at the recommended imaging timrather than immediately postimplantation, may involve adtional catheterization. It is hoped that a more convenicontrast-enhancing technique will become available innear future.

I. Training requirement for physics personnel

For a member of the physics staff to perform independwork in permanent prostate brachytherapy, it isrecom-mendedthat a minimum of five documented cases be pformed under the direct supervision of an experienced phcist. In this context, an experienced physicist is one~a! whosatisfies the above training requirement, or~b! who has per-formed a minimum of 20 documented cases independen

J. Recommendations regarding commercial treatmentplanning systems

It is recommendedthat the TG43 dose calculation formaism beexplicitly represented in commercial treatment planing systems for prostate seed implantation. This impliesthe half-life, the dose rate constant, the radial dose funct


g

-

ee

f

u-rn

f-s-of

-

ey

lyralu-

du-,

i-te

t

-i-

.

-atn,

and as a minimum the anisotropy constant are separaspecified in the source data library. Thus if future changare to occur on any of the parameters for a given radioacsource, they can be easily and uniformly updated by the uof the system with the least confusion.

It is recommendedthat software facilities be implementeto generate the DVH for the target volume and the DSHthe rectal wall, both in preoperative planning and for postiplant evaluation.

In addition to these practical guidelines, the medicphysicist should observe the recommendations of AAPTask Group No. 56 with regard to the code of practicebrachytherapy physics,42 and of AAPM Task Group No. 40with regard to quality assurance for radiation oncologygeneral.108

V. ISSUES FOR FUTURE CONSIDERATION

This section contains a discussion of some dosimetricradiobiologic effects on which consensus understanding dnot yet exist among investigators of prostate brachytheraIn addition, areas of current and future investigation are idtified to aid practising medical physicists in designing thclinical seed implant programs.

A. Anisotropic dose calculation

The AAPM TG43 report41 contains extensive tabulatioof the anisotropy functions for125I and103Pd single seeds. Inprinciple, it is not difficult to incorporate the anisotropfunction formalism at the planning stage. However, indoing one needs to make certain assumptions about theentation of the radioactive seeds, e.g., being perfectly aligalong the needle insertion direction. Such an assumptioprobably quite valid in the case of seed strands compareloose seeds, but for any given case it is impossible to prethe extent and direction of splaying that will occur. On tother hand, the anisotropy constant is an averaged quaweighted by the solid angle, and therefore represents theestimate of the dose surrounding a radioactive seed of interminate orientation. Use of the anisotropy function formism in postimplant dosimetry is technically more difficusince the orientation of each seed must be determinedlocating both ends of the seed. Until automated seed recstruction software becomes widely available, the posource approximation appears to be the more appropriatemalism.

The dosimetric effects of anisotropy for125I were dis-cussed by Linget al.109 There appear to be rather large dferences in the dose distribution and the mPD betweenanisotropy function formalism and the point source appromation widely used at present. To maintain uniform standof dose reporting in prostate seed implant, investigatorsurged to document the dose calculation formalism in thplanning and/or postimplant dosimetry procedures.

B. Interseed effect

Mutual attenuation by neighboring seeds has beenported to be significant.110,111 Meigooni et al.111 performed

fffe

iniehedoce

oedith

th(

-o

thhea1

enanssa

aeetaltio

oduaath

ley-o

o-

itydo-

nt

lantrs

thethatelyof

ng

ros-

forme-

celles,

rcht-

ng.

BE

seat

tont,ity.

ofn. Aongntndteduc-red

assenio-di-ngptf


Solid Water measurement and Monte Carlo calculationsexamine the dose perturbation in a two-plane implant o33 seed arrays. They concluded that the inter-seed ewould reduce the peripheral dose by 6% for125I seeds. Whilethe dosimetric impact of the inter-seed effect may be of clcal concern in simple, regular configurations, the overallfect in prostate implants is not clear. In practice, only tnearest-neighbor seeds are likely to produce appreciableperturbation. The solid angle sustained by a seed at 1average distance is sufficiently small that the volume of pturbation is for most purposes negligible.

C. Tissue heterogeneity

The major cause of tissue heterogeneity is calcified depits in the prostate gland, which occur in a small percentagpatients. The calcification presents on TRUS and CT stuas hyperechoic and high density regions, in contrast tosurrounding fibromuscular tissue. In the energy range of125Iand103Pd radionuclides, where the photoelectric effect isdominant absorption process, the presence of calciumZ520) in fibromuscular tissue (Z57.6) leads to three dosimetric effects:~a! the dose rate constant is different in the twmedia, resulting in different absorbed dose;~b! the radialdose function is modified by the increased attenuation ofhigh Z material;~c! increased dose deposition occurs in tsoft tissue at the interface of heterogeneity, due to a grenumber of photoelectrons, which have a range of aboutm.112–117The overall dosimetric effect depends on the extand the microscopic structure of calcification and the implconfiguration. As a first approximation, the ratio of maenergy attenuation coefficients of calcium to muscle is 2430 keV, and 23 at 20 keV. At present, there is no clinicstudy to gauge the actual impact of such tissue heterogenIt is prudent to identify patients who present with tissue herogeneity under planning radiological studies, and to evate the efficacy and the optimal strategy of seed implantaon an individual basis.

The same physical principles may lead to variabilitydose deposition in malignant versus normal histologiesto different elemental compositions. It is not yet clewhether the physical laws translate to a therapeutic advtage for adenocarcinoma of the prostate, and if so, whatmagnitude of such advantage is.

D. Biological models

The linear-quadratic cell-kill model was extended by Dato take into account~a! the decaying dose rate in brachtherapy,~b! dose rate difference across dosimetric inhomgeneity,~c! tumor cell proliferation, and~d! repair of suble-thal damage for low dose-rate radiation.118,119 This modelwas used by Linget al. to examine the effect of dose hetergeneity in prostate seed implants.96 The authors concludedthat there might be some advantage in dose heterogeneabout 20% above the prescribed dose, but beyond that,would be ‘‘wasted’’ in terms of producing cell kill. The biologically effective dose~BED! and cell surviving fractionspredicted based on the model have been used as endpoi


to3ct

-f-

semr-

s-ofese

e

e

ter0tt

tlity.-u-n

fe

rn-e

-

atse

s to

compare alternative treatment plans in prostate impoptimization.92 Assumptions on the following parametemust be made to apply the Dale model: thea-to-b ratio, thevalue ofa, the potential doubling time for tumor cells (Tpot),and the mean time for repair of sublethal damage. Givenuncertainties in these parameters, it must be concludedthe biological models should not be taken as quantitativpredictive, but rather as a guide of the relative efficacycompeting treatment plans. In addition to the cell survivifraction, the tumor control probability~TCP! can be calcu-lated based on the BED with additional assumptions on ptate tumor dose response data.96,120

The commonly quoted prescription dose of 115 Gy103Pd implants is the dose estimated to have the same ‘‘tidose-factor’’~TDF!121 as that corresponding to 145 Gy~con-verted from 160 Gy of pre-TG43 dose! from 125I implants.Using the linear quadratic model to compare the relativekill effectiveness of the two radionuclides for these dosLing122 has shown that103Pd may be more effective forTpot

of a few days and that125I may be more effective for longeTpot. The determination of the most efficacious dose for eatype of implants involves ongoing analysis of clinical oucome, dosimetric specification and radiobiological modeli

E. Relative biological effectiveness

The relative biological effectiveness~RBE! was measuredby Ling et al.123 for 125I and 103Pd and by Freemanet al.124

and Marcheseet al.125,126 for 125I. Relative to 60Co and atdose rates relevant to permanent prostate implant, the Rwas reported to be about 1.4 for125I and about 1.9 for103Pd.123 The enhanced cell inactivation for a given doreflects the additional biological effects of the radiation thare not described by the physical quantities.

These radiobiological effects are currently not taken inaccount in the clinical dosimetry for prostate seed implaand are considered theoretical advantages of this modal

F. Time course of target volume change

Work is continuing on characterizing the time courseprostatic volume change subsequent to seed implantatiomore complete understanding of this issue will have a strimpact on optimal dosimetric planning and postimplaanalysis. Any significant differences in the pattern of glaswelling and resolution may be dosimetrically compensafor in planning for the specific radionuclide used, thus reding the variance in the effective treatment dose deliveacross the patient population.

G. Differential dose planning and delivery

The notion of planning the dose distribution to encompthe primary foci of the tumor in a high dose region is oftan attractive one in treatment planning. It is justified radbiologically on the basis that higher dose is required to eracate higher tumor cell density. Advances in tumor imagifor prostate cancer will lend more credibility to the conceof differential interstitial irradiation. It is the nature o

thrmzehithnbraou

eun

edeer

ot

nlon

cajog-o

fooimluiorlimido

s-e

Rin-

atiy.

nry

n

rol.,

for

ns-n-

ane-l ul-.

anem.

ns-entative

S.m-

-tate

.’’

cerlan-

nal.

ti-ars

.n-erm

.

iat.

nts

icd

ofntsd.

emaech-

mInt.

ific


brachytherapy to accept certain high dose regions intreatment volume in order to achieve reasonably unifodose coverage at a lower isodose level. With computerioptimization strategies, it is possible to routinely achieve tgoal by placing the high dose volume at the focal area ofgross tumor. If the goal is achieved without sacrificing aother aspects of the clinical treatment plan, then it mayhypothesized that differential dose planning offers a thepeutic advantage compared to dosimetric planning withregard to the locations of tumor foci.

H. Intraoperative seed localization and dosimetry

While postimplant dosimetry is important for quantitativevaluation of dosimetric outcome, prostate brachytherapytimately will benefit from intraoperative seed localizatiofollowed by real-time computerized dosimetry, all performwith the patient still under anesthesia and in the treatmposition. Thus any significant underdosage can be discovand remedied by additional implantation before the endthe procedure. There is then a reasonable expectationevery implant will deliver the intended dose, where the odosimetric variability is due to ongoing edematous reacti

I. Correlation of dosimetric and clinical outcomes

A number of studies21,33,38,71suggest that the dosimetrioutcome of a prostate seed implant ultimately plays a mrole in predicting the likelihood of local relapse and/or lonterm treatment-related morbidity. The natural progressionthe disease is such that extensive follow-up is requiredclinical outcome analysis. Such clinical correlation with dsimetric predictors will be aided by more consistent postplant analysis and quantitative, organ-specific dose evation on a larger scale. Careful treatment plan optimizatand dosimetric outcome analysis will in turn provide an eaindication of treatment effectiveness for a prostate seedplant. Such is the goal of the present effort in seeking wsuccess of interstitial implantation in the managementprostate cancer.

ACKNOWLEDGMENTS

We thank W. S. Bice, W. M. Butler, S. Pai, B. R. Pretidge, and R. G. Stock for helpful discussions on the contof this report and various aspects of the current practiceprostate brachytherapy. We also thank members of thediation Therapy Committee of the AAPM chaired by Jatder Palta for careful review of the report.

a!This report represents the recommendations of the American Associof Physicists in Medicine~AAPM! for permanent prostate brachytherap

b!Electronic mail: [email protected]. V. D’Amico and C. N. Coleman, ‘‘Role of interstitial radiotherapy ithe management of clinically organ-confined prostate cancer: The justill out,’’ J. Clin. Oncol.14, 304–315~1996!.

2W. F. Whitmore, B. Hilaris, and H. Grabstald, ‘‘Retropubic implantatioto iodine-125 in the treatment of prostatic cancer,’’ J. Urol.~Baltimore!108, 918–920~1972!.

3L. L. Anderson, ‘‘Spacing nomograph for interstitial implants of125Iseeds,’’ Med. Phys.3, 48–51~1976!.


e

dse

ye-t

l-

ntedf

haty.

r

fr

--a-ny

-ef

ntina-

on

is

4L. L. Anderson, J. V. Moni, and L. B. Harrison, ‘‘A nomograph fopermanent implants of palladium-103 seeds,’’ Int. J. Radiat. OncBiol., Phys.27, 129–136~1993!.

5U. K. Henschke and P. Ceve, ‘‘Dimension averaging a simple methoddosimetry of interstitial implants,’’ Radiobiol. Radiother.~Berl! 9, 287–298 ~1968!.

6B. S. Hilaris, D. Nori, and L. L. Anderson,Atlas of Brachytherapy~Mac-millan, New York, 1988!.

7H. H. Holm, N. Juul, J. F. Pedersen, H. Hansen, and I. Stroyer, ‘‘Traperineal125iodine seed implantation in prostatic cancer guided by trasrectal ultrasonography,’’ J. Urol.~Baltimore! 130, 283–286~1983!.

8J. C. Blasko, H. Ragde, and D. Schumacher, ‘‘Transperineal percutous iodine-125 implantation for prostatic carcinoma using transrectatrasound and template guidance,’’ Endocurie./Hypertherm. Oncol3,131–139~1987!.

9J. C. Blasko, P. D. Grimm, and H. Ragde, ‘‘Brachytherapy and orgpreservation in the management of carcinoma of the prostate,’’ SRadiat. Oncol.3, 240–249~1993!.

10P. D. Grimm, J. C. Blasko, and H. Radge, ‘‘Ultrasound-guided traperineal implantation of iodine-125 and palladium-103 for the treatmof early stage prostate cancer; Technical concepts in planning, opertechnique and evaluation,’’ Atlas Urol. Clin. North Am.2, 113–125~1994!

11K. Wallner, S. T. Chiu-Tsao, J. Roy, V. E. Arterbery, W. Whitmore,Jain, B. Minsky, P. Russo, and Z. Fuks, ‘‘An improved method for coputerized tomography-planned transperineal125iodine prostate implants,’’J. Urol. ~Baltimore! 146, 90–95~1991!.

12K. Wallner, J. Roy, M. Zelefsky, Z. Fuks, and L. Harrison, ‘‘Fluoroscopic visualization of the prostatic urethra to guide transperineal prosimplantation,’’ Int. J. Radiat. Oncol., Biol., Phys.29, 863–867~1994!.

13W. F. Whitmore, B. Hilaris, M. Batata, P. Sogani, H. Herr, and MMorse, ‘‘Interstitial radiation: short-term palliation or curative therapy?Urol. 25, 24–29~1985!.

14J. D. Morton, L. B. Harrison, and R. E. Peschel, ‘‘Prostatic cantherapy: Comparison of external-beam radiation and I-125 seed imptation treatment of stagesB andC neoplasms,’’ Radiology159, 249–252~1986!.

15J. D. Morton and R. E. Peschel, ‘‘Iodine-125 implants versus exterbeam therapy for stagesA2, B, and C prostate cancer,’’ Int. J. RadiatOncol., Biol., Phys.14, 1153–1157~1988!.

16D. A. Kuban, A. M. el-Mahdi, and P. F. Schellhammer, ‘‘I-125 interstial implantation for prostate cancer. What have we learned 10 yelater?’’ Cancer~N.Y.! 63, 2415–2420~1989!.

17Z. Fuks, S. A. Leibel, K. E. Wallner, C. B. Begg, W. R. Fair, L. LAnderson, B. S. Hilaris, and W. F. Whitmore, ‘‘The effect of local cotrol on metastatic dissemination in carcinoma of the prostate: long-tresults in patients treated with125I implantation,’’ Int. J. Radiat. Oncol.,Biol., Phys.21, 537–547~1991!.

18C. D. Koprowski, K. G. Berkenstock, A. M. Borofski, J. C. Ziegler, D. ALightfoot, and L. W. Brady, ‘‘External beam irradiation versus125iodineimplant in the definitive treatment of prostate carcinoma,’’ Int. J. RadOncol., Biol., Phys.21, 955–960~1991!.

19T. P. Weyrich, S. J. Kandzari, and P. R. Jain, ‘‘Iodine-125 seed implafor prostatic carcinoma. Five- and ten-year follow-up,’’ Urol.41, 122–126 ~1993!.

20M. J. Zelefsky and W. F. Whitmore, Jr., ‘‘Long-term results of retropubpermanent125iodine implantation of the prostate for clinically localizeprostatic cancer,’’ J. Urol.~Baltimore! 158, 23–29~1997!.

21R. Nath, K. Roberts, M. Ng, R. Peschel, and Z. Chen, ‘‘Correlationmedical dosimetry quality indicators to the local tumor control in patiewith prostate cancer treated with iodine-125 interstitial implants,’’ MePhys.25, 2293–2307~1998!.

22P. L. Vijverberg, K. H. Kurth, L. E. Blank, N. F. Dabhoiwala, T. H. dReijke, and K. Koedooder, ‘‘Treatment of localized prostatic carcinousing the transrectal ultrasound guided transperineal implantation tnique,’’ Eur. Urol.21, 35–41~1992!.

23K. Wallner, J. Roy, M. Zelefsky, Z. Fuks, and L. Harrison, ‘‘Short-terfreedom from disease progression after I-125 prostate implantation,’’J. Radiat. Oncol., Biol., Phys.30, 405–409~1994!.

24J. C. Blasko, K. Wallner, P. D. Grimm, and H. Ragde, ‘‘Prostate specantigen based disease control following ultrasound guided125iodine im-plantation for stageT1/T2 prostatic carcinoma,’’ J. Urol.~Baltimore!154, 1096–1099~1995!.

ty

any

hy

, atiool

ivas.

. Conate

sisou

, Fdet o

tra-pl

.y-ed

t o

tyia

s,fol.,

ero

eate

ngt.

aneniol

ndc-

3,’

ll-theys

er

aneia

atd

ee-

‘‘Ae-ol.,

ndt. J.

-

-

‘‘Ab-

nt,’’

-

M.ate

teol.,

edfluo-

i-cifi-

,

W.os-

gim-

n,te

tedost-J.

ofrma-iol.,

r-ca-

ndedim-

k,put-py’’

lin,gis-iat.

. J.ra-


25K. Wallner, J. Roy, and L. Harrison, ‘‘Tumor control and morbidifollowing transperineal iodine-125 implantation for stageT1/T2 prostaticcarcinoma,’’ J. Clin. Oncol.14, 449–453~1996!.

26B. R. Prestidge, D. C. Hoak, P. D. Grimm, H. Ragde, W. Cavanagh,J. C. Blasko, ‘‘Posttreatment biopsy results following interstitial brachtherapy in early-stage prostate cancer,’’ Int. J. Radiat. Oncol., Biol., P37, 31–39~1997!.

27S. H. Stokes, J. D. Real, P. W. Adams, J. C. Clements, S. WuertzerW. Kan, ‘‘Transperineal ultrasound-guided radioactive seed implantafor organ-confined carcinoma of the prostate,’’ Int. J. Radiat. OncBiol., Phys.37, 337–341~1997!.

28D. C. Beyer and J. B. Priestley, Jr., ‘‘Biochemical disease-free survfollowing 125I prostate implantation,’’ Int. J. Radiat. Oncol., Biol., Phy37, 559–563~1997!.

29H. Ragde, J. C. Blasko, P. D. Grimm, G. M. Kenny, J. E. Sylvester, DHoak, K. Landin, and W. Cavanagh, ‘‘Interstitial iodine-125 radiatiwithout adjuvant therapy in the treatment of clinically localized prostcarcinoma,’’ Cancer~N.Y.! 80, 442–453~1997!.

30K. W. Kaye, D. J. Olson, and J. T. Payne, ‘‘Detailed preliminary analyof 125iodine implantation for localized prostate cancer using percutaneapproach,’’ J. Urol.~Baltimore! 153, 1020–1025~1995!.

31P. Iversen, M. Bak, N. Juul, F. Laursen, H. von der Maase, L. NielsenRasmussen, S. Torp-Pedersen, and H. H. Holm, ‘‘Ultrasonically gui125iodine seed implantation with external radiation in managemenlocalized prostatic carcinoma,’’ Urol.34, 181–186~1989!.

32P. Iversen, F. Rasmussen, and H. H. Holm, ‘‘Long-term results of ulsonically guided implantation of125I seeds combined with external irradiation in localized prostatic cancer,’’ Scand. J. Urol. Nephrol. Sup138, 109–115~1991!.

33P. L. Vijverberg, L. E. Blank, N. F. Dabhoiwala, T. M. de Reijke, CKoedooder, A. A. Hart, K. H. Kurth, and D. Gonzalez Gonzalez, ‘‘Analsis of biopsy findings and implant quality following ultrasonically-guid125I implantation for localized prostatic carcinoma,’’ Br. J. Urol.72, 470–477 ~1993!.

34J. B. Priestly and D. C. Beyer, ‘‘Guided brachytherapy for treatmenconfined prostate cancer,’’ Urol.40, 27–32~1992!.

35V. E. Arterbery, K. Wallner, J. Roy, and Z. Fuks, ‘‘Short-term morbidifrom CT-planned transperineal I-125 prostate implants,’’ Int. J. RadOncol., Biol., Phys.25, 661–667~1993!.

36L. Kleinberg, K. Wallner, J. Roy, M. Zelefsky, V. E. Arterbery, Z. Fukand L. Harrison, ‘‘Treatment-related symptoms during the first yearlowing transperineal125I prostate implantation,’’ Int. J. Radiat. OncolBiol., Phys.28, 985–990~1994!.

37S. Nag, D. D. Scaperoth, R. Badalament, S. A. Hall, and J. Burg‘‘Transperineal palladium-103 prostate brachytherapy: analysis of mbidity and seed migration,’’ Urol.45, 87–92~1995!.

38K. Wallner, J. Roy, and L. Harrison, ‘‘Dosimetry guidelines to minimizurethral and rectal morbidity following transperineal I-125 prostbrachytherapy,’’ Int. J. Radiat. Oncol., Biol., Phys.32, 465–471~1995!.

39R. G. Stock, N. N. Stone, and C. Iannuzzi, ‘‘Sexual potency followiinteractive ultrasound-guided brachytherapy for prostate cancer,’’ InRadiat. Oncol., Biol., Phys.35, 267–272~1996!.

40J. J. Prete, B. R. Prestidge, W. S. Bice, J. L. Friedland, R. G. Stock,P. D. Grimm, ‘‘A survey of physics and dosimetry practice of permanprostate brachytherapy in the United States,’’ Int. J. Radiat. Oncol., BPhys.40, 1001–1005~1998!.

41R. Nath, L. L. Anderson, G. Luxton, K. A. Weaver, J. F. Williamson, aA. S. Meigooni, ‘‘Dosimetry of interstitial brachytherapy sources: Reommendations of the AAPM Radiation Committee Task Group No. 4Med. Phys.22, 209–234~1995!.

42R. Nath, L. L. Anderson, J. Meli, A. J. Olch, J. A. Stitt, and J. F. Wiiamson, ‘‘Code of practice for brachytherapy physics: Report ofAAPM Radiation Therapy Committee Task Group No. 56,’’ Med. Ph24, 1557–1598~1997!.

43J. Desai, R. G. Stock, N. N. Stone, C. Iannuzzi, and J. K. DeWynga‘‘Acute urinary morbidity following I-125 interstitial implantation of theprostate gland,’’ Radiat. Oncol. Invest.6, 135–141~1998!.

44E. M. Messing, J. B. Zhang, D. J. Rubens, R. A. Brasacchio, J. G. StrA. Soni, P. G. Okunieff, and Y. Yu, ‘‘Intraoperative optimized inversplanning for prostate brachytherapy: Early experience,’’ Int. J. RadOncol., Biol., Phys.44, 801–808~1999!.

45N. N. Stone, R. G. Stock, J. K. DeWyngaert, and A. Tabert, ‘‘Prostbrachytherapy: Improvements in prostate volume measurements and


d-s.

ndn.,

l

.

s

.df

-

.

f

t.

-

s,r-

J.

dt.,

’

.

t,

g,

t.

eose

distribution using interactive ultrasound guided implantation and thrdimensional dosimetry,’’ Radiat. Oncol. Invest.3, 185–195~1995!.

46R. G. Stock, N. N. Stone, M. F. Wesson, and J. K. DeWyngaert,modified technique allowing interactive ultrasound-guided thredimensional transperineal prostate implantation,’’ Int. J. Radiat. OncBiol., Phys.32, 219–225~1995!.

47B. R. Prestidge, E. B. Butler, D. Shaw, and Van. McComas, ‘‘Ultrasouguided placement of transperineal prostatic afterloading catheters,’’ InRadiat. Oncol., Biol., Phys.28, 263–266~1994!.

48Interstitial Collaborative Working Group (ICWG), Interstitial Brachytherapy: Physical, Biological, and Clinical Considerations, edited by L.L. Andersonet al. ~Raven, New York, 1990!.

49D. E. Mellenberg and R. W. Kline, ‘‘Verification of manufacturersupplied125I and 103Pd air-kerma strengths,’’ Med. Phys.22, 1495–1497~1995!.

50V. Feygelman, B. K. Noriega, R. M. Sanders, and J. L. Friedland,simple method for verifying activity of iodine-125 seeds in rigid absorable suture,’’ Med. Dosim.21, 261–262~1996!.

51W. M. Butler, A. T. Dorsey, K. R. Nelson, G. M. Merrick, ‘‘Qualityassurance calibration of I-125 Rapid Strand in a sterile environmeInt. J. Radiat. Oncol., Biol., Phys.41, 217–222~1998!.

52W. M. Butler and G. S. Merrick, ‘‘I-125 Rapid Strand loading technique,’’ Radiat. Oncol. Invest.4, 48–49~1996!.

53V. Feygelman, J. L. Friedland, R. M. Sanders, B. K. Noriega, and J.Pow-Sang, ‘‘Improvement in dosimetry of ultrasound-guided prostimplants with the use of multiple stabilization needles,’’ Med. Dosim.21,109–112~1996!.

54M. Dattoli and K. Wallner, ‘‘A simple method to stabilize the prostaduring transperineal prostate brachytherapy,’’ Int. J. Radiat. OncBiol., Phys.38, 341–342~1997!.

55K. W. Keye, D. J. Olsen, D. J. Lightner, and J. T. Payne, ‘‘Improvtechnique for prostate seed implantation: combined ultrasound androscopic guidance,’’ J. Endourology6, 61–66~1992!.

56Y. Yu, F. M. Waterman, N. Suntharalingam, and A. Schulsinger, ‘‘Limtations of the minimum peripheral dose as a parameter for dose specation in permanent125I prostate implants,’’ Int. J. Radiat. Oncol., Biol.Phys.34, 717–725~1996!.

57P. L. Roberson, V. Narayana, D. L. McShan, R. J. Winfield, and P.McLaughlin, ‘‘Source placement error for permanent implant of the prtate,’’ Med. Phys.24, 251–257~1997!.

58J. N. Roy, C. C. Ling, K. E. Wallner, and L. L. Anderson, ‘‘Determininsource strength and source distribution for a transperineal prostateplant,’’ Endocurietherapy/Hyperthermia Oncol.12, 35–41~1996!.

59F. M. Waterman, N. Yue, S. Reisinger, A. Dicker, and B. W. Cor‘‘Effect of edema on the post-implant dosimetry of an I-125 prostaimplant: A case study,’’ Int. J. Radiat. Oncol., Biol., Phys.38, 335–339~1997!.

60F. M. Waterman, N. Yue, B. W. Corn and A. Dicker, ‘‘Edema associawith I-125 or Pd-103 prostate brachytherapy and its impact on pimplant dosimetry: An analysis based on serial CT acquisition,’’ Int.Radiat. Oncol., Biol., Phys.41, 1069–1077~1998!.

61B. R. Prestidge, W. S. Bice, E. J. Kiefer, and J. J. Prete, ‘‘Timingcomputed tomography-based postimplant assessment following penent transperineal prostate brachytherapy,’’ Int. J. Radiat. Oncol., BPhys.40, 1111–1115~1998!.

62J. N. Roy, K. E. Wallner, P. J. Harrington, C. C. Ling, and L. L. Andeson, ‘‘A CT-based evaluation method for permanent implants: Applition to prostate,’’ Int. J. Radiat. Oncol., Biol., Phys.26, 163–169~1993!.

63V. Narayana, P. L. Roberson, A. T. Pu, H. Sandler, R. H. Winfield, aP. W. McLaughlin, ‘‘Impact of differences in ultrasound and computtomography volumes on treatment planning of permanent prostateplants,’’ Int. J. Radiat. Oncol., Biol., Phys.37, 1181–1185~1997!.

64M. Roach, P. Faillace-Akazawa, C. Malfatti, J. Holland, and H. Hrica‘‘Prostate volumes defined by magnetic resonance imaging and comerized tomographic scans for three-dimensional conformal radiotheraInt. J. Radiat. Oncol., Biol., Phys.35, 1011–1018~1996!.

65V. Narayana, P. L. Roberson, R. J. Winfield, and P. W. McLaugh‘‘Impact of ultrasound and computed tomography prostate volume retration on evaluation of permanent prostate implants,’’ Int. J. RadOncol., Biol., Phys.39, 341–346~1997!.

66M. A. Moerland, H. K. Wijrdeman, R. Beersma, C. J. Bakker, and JBattermann, ‘‘Evaluation of permanent I-125 prostate implants using

ol

. JtatRa

25

ss,’’

anzey-

ern

ld,t o

H

.

-

r-oree

s-.

hee

onol.

ca43

-

ottienesrg

ts

fo

of

in

ofund

f

-

.ns-queol.,

g

n-to

rinte,’’

rsti-

C.ing

W.s-

ntiat.

, J.ingol.,

h,ys.

-

a-

n

l-M

n,andf

in.

su-

for

etry,’’

nt.

,’’

or


diography and magnetic resonance imaging,’’ Int. J. Radiat. OncBiol., Phys.37, 927–933~1997!.

67D. F. Dubois, B. R. Prestidge, L. A. Hotchkiss, W. S. Bice, and JPrete, ‘‘Source localization following permanent transperineal prosinterstitial brachytherapy using magnetic resonance imaging,’’ Int. J.diat. Oncol., Biol., Phys.39, 1037–1041~1997!.

68J. Willins and K. Wallner, ‘‘CT-based dosimetry for transperineal I-1prostate brachytherapy,’’ Int. J. Radiat. Oncol., Biol., Phys.39, 347–353~1997!.

69W. S. Bice and B. R. Prestidge, ‘‘A review of postimplant quality assement in permanent transperineal interstitial prostate brachytherapyBrachytherapy Int.13, 297–313~1997!.

70W. S. Bice, B. R. Prestidge, P. D. Grimm, J. L. Friedland, V. FeygelmM. Roach, J. J. Prete, D. F. Dubois, and J. C. Blasko, ‘‘Centralimulti-institutional post-implant analysis for interstitial prostate brachtherapy,’’ Int. J. Radiat. Oncol., Biol., Phys.41, 921–927~1998!.

71R. G. Stock, N. N. Stone, A. Tabert, C. Iannuzzi, and J. K. DeWynga‘‘A dose-response study for I-125 prostate implants,’’ Int. J. Radiat. Ocol., Biol., Phys.41, 101–108~1998!.

72B. S. Hilaris, W. F. Whitmore, Jr., M. A. Batata, and H. Grabsta‘‘Radiation therapy and pelvic node dissection in the managemencancer of the prostate,’’ Am. J. Radiol.121, 832–838~1974!.

73O. Glasser, E. H. Quimby, L. S. Taylor, J. L. Weatherwax, and R.Morgan,Physical Foundations of Radiology, 3rd ed.~Harper and Row,New York, 1961!.

74Radium Dosage—the Manchester System, 2nd ed., edited by W. JMeredith ~E. and S Livingstone, Edinburgh, 1967!.

75L. L. Anderson, inHandbook of Interstitial Brachytherapy, edited by B.S. Hilaris ~Publishing Sciences Group, Acton, MA, 1975!.

76L. L. Anderson, H. M. Kuan, and I. Y. Ding, inModern Interstitial andIntracavitary Radiation Cancer Management, edited by F. W. George~Masson, New York, 1981!, pp. 9–15.

77L. L. Anderson and I. Y. Ding,Dosimetric Considerations for Iodine125, in Afterloading: 20 Years of Experience~Memorial Sloan-KetteringCancer Center, New York, 1975!, pp. 63–72.

78B. S. Hilaris, J. G. Holt, and J. St. Germain,The Use of Iodine-125 forInterstitial Implants, DHEW Publication~FDA! 76-8022~U.S. Depart-ment of Health, Education and Welfare, Rockville, MD, 1975!.

79V. Krishnaswamy, ‘‘Dose distribution around an125I seed source in tis-sue,’’ Radiology126, 489–491~1978!.

80H. D. Kubo, W. F. Hanson, J. F. Williamson, R. W. Kline, B. M. Cousey, S. M. Seltzer, R. E. Shuping, L. DeWerd, and M. Langton, ‘‘Repof the Ad Hoc Committee of the AAPM Radiation Therapy Committon I-125 sealed source dosimetry,’’ Int. J. Radiat. Oncol., Biol., Phys.40,697–702~1998!.

81C. C. Ling, D. York, I. J. Spiro, D. Kubiatowicz, and D. Bennett, ‘‘Phyics dosimetry of125I seeds of a new design for interstitial implant,’’ Int. JRadiat. Oncol., Biol., Phys.9, 1747–1752~1983!.

82R. W. Luse, J. Blasko, and P. Grimm, ‘‘A method for implementing tAmerican Association of Physicists in Medicine Task Group-43 dosimtry recommendations for125I transperineal prostate seed implantscommercial treatment planning systems,’’ Int. J. Radiat. Oncol., BiPhys.37, 737–741~1997!.

83W. S. Bice, B. R. Prestidge, J. J. Prete, and D. F. Dubois, ‘‘Cliniimpact of implementing the recommendations of AAPM Task Grouppermanent prostate brachytherapy using125I, ’’ Int. J. Radiat. Oncol.,Biol., Phys.40, 1237–1241~1998!.

84R. Loevinger, ‘‘Wide-angle free-air chamber for calibration of lowenergy brachytherapy sources,’’ Med. Phys.20, 907 ~1993!.

85J. F. Williamson, B. M. Coursey, W. F. Hanson, R. Nath, and G. Ibb‘‘Guidance to users of Nycomed Amersham and North American Sctific, Inc., I-125 Interstitial Sources: Dosimetry and calibration changRadiation Therapy Committee Ad Hoc Subcommittee on Low-EneSeed Dosimetry,’’ Med. Phys.26, 570–573~1998!.

86V. Krishnaswamy, ‘‘Dose tables for125I seed implants,’’ Radiology132,729–730~1979!.

87G. U. V. Rao, P. T. Kan, and R. Howells, ‘‘Interstitial volume implanwith I-125 seeds,’’ Int. J. Radiat. Oncol., Biol., Phys.7, 431–438~1991!.

88F. W. Waterman and K. A. Strubler, ‘‘Absorbed dose determinationinterstitial 125I boost therapy,’’ Med. Phys.10, 155–158~1983!.

89A. Wu, R. D. Zwicker, and E. S. Sternick, ‘‘Tumor dose specificationI-125 seed implants,’’ Med. Phys.12, 27–31~1985!.

90Y. Yu, ‘‘A non-divergent specification of the mean treatment dose


.,

.e-

-J.

,d

t,-

f

.

t

-

,

l

,-:y

r

interstitial brachytherapy,’’ Med. Phys.23, 905–909~1996!.91J. E. Dawson, T. Wu, T. Roy, J. Y. Gu, and H. Kim, ‘‘Dose effects

seed placement deviations from pre-planned positions in ultrasoguided prostate implants,’’ Radiother. Oncol.32, 268–270~1994!.

92Y. Yu and M. C. Schell, ‘‘A genetic algorithm for the optimization oprostate implants,’’ Med. Phys.23, 2085–2091~1996!.

93L. L. Anderson, ‘‘A ‘natural’ volume-dose histogram for brachytherapy,’’ Med. Phys.13, 898–903~1986!.

94A. van’t Riet, H. J. te Loo, A. F. Ypma, A. C. Mak, F. H. van Slooten, CJ. Hoekstra, and M. C. Stenfert Kroese, ‘‘Ultrasonically guided traperineal seed implantation of the prostate: modification of the techniand qualitative assessment of implants,’’ Int. J. Radiat. Oncol., BiPhys.24, 555–558~1992!.

95R. Nath, A. S. Meigooni, and A. Melillo, ‘‘Some treatment planninconsiderations for103Pd and125I permanent interstitial implants,’’ Int. J.Radiat. Oncol., Biol., Phys.28, 971–977~1994!.

96C. C. Ling, J. Roy, N. Sahoo, K. Wallner, and L. L. Anderson, ‘‘Quatifying the effect of dose inhomogeneity in brachytherapy: Applicationpermanent prostatic implant with125I seeds,’’ Int. J. Radiat. Oncol., Biol.,Phys.28, 971–978~1994!.

97A. van’t Riet, A. C. Mak, M. A. Moerland, L. H. Elders, and W. van deZee, ‘‘A conformation number to quantify the degree of conformalitybrachytherapy and external beam irradiation: application to the prostaInt. J. Radiat. Oncol., Biol., Phys.37, 731–736~1997!.

98C. B. Saw and N. Suntharalingam, ‘‘Quantitative assessment of intetial implants,’’ Int. J. Radiat. Oncol., Biol., Phys.20, 135–139~1991!.

99J. N. Roy, K. E. Wallner, S. T. Chiu-Tsao, L. L. Anderson, and C.Ling, ‘‘CT-based optimized planning for transperineal prostate plannwith customized template,’’ Int. J. Radiat. Oncol., Biol., Phys.21, 483–489 ~1991!.

100V. Narayana, P. L. Roberson, R. J. Winfield, M. L. Kessler, and P.McLaughlin, ‘‘Optimal placement of radioisotopes for permanent protate implants,’’ Radiology199, 457–460~1996!.

101J. Pouliot, D. Tremblay, J. Ro, and S. Filice, ‘‘Optimization of permane125I prostate implants using fast simulated annealing,’’ Int. J. RadOncol., Biol., Phys.36, 711–720~1996!.

102Y. Yu, JB Y. Zhang, R. A. Brasacchio, P. G. Okunieff, D. J. RubensG. Strang, A. Soni, and E. M. Messing, ‘‘Automated treatment plannengine for prostate seed implant brachytherapy,’’ Int. J. Radiat. OncBiol., Phys.43, 647–652~1999!.

103Y. Chen, R. E. Stanton, R. J. Holst, C. D. Koprowski, and E. B. Krisc‘‘Treatment planning for prostate implant with loose seeds,’’ Med. Ph24, 1141–1145~1997!.

104S. Kirkpatrick, C. D. Gelatt, and M. P. Vecci, ‘‘Optimization by simulated annealing,’’ Science220, 671–680~1983!.

105D. E. Goldberg,Genetic Algorithms in Search, Optimization and Mchine Learning~Addison-Wesley, Reading, MA, 1989!.

106Y. Yu, ‘‘Multiobjective decision theory for computational optimizatioin radiation therapy,’’ Med. Phys.24, 1445–1454~1997!.

107M. M. Goodsitt, P. L. Carson, S. Witt, D. L. Hykes, J. M. Kofler, ‘‘Reatime B-mode ultrasound quality control test procedures: Report of AAPUltrasound Task Group No. 1,’’ Med. Phys.25, 1385–1406~1998!.

108G. J. Kutcher, L. Coia, M. Gillin, W. F. Hanson, S. Leibel, R. J. MortoJ. R. Palta, J. A. Purdy, L. E. Reinstein, G. K. Svensson, M. Weller,L. Wingfield, ‘‘Comprehensive QA for radiation oncology: Report oAAPM Radiation Therapy Committee Task Group 40,’’ Med. Phys.21,581–618~1994!.

109C. C. Ling, L. L. Anderson, and W. U. Shipley, ‘‘Dose inhomogeneityinterstitial implants using125I seeds,’’ Int. J. Radiat. Oncol., Biol., Phys5, 419–425~1979!.

110G. S. Burns and D. E. Raeside, ‘‘The accuracy of single-seed doseperposition for I-125 implants,’’ Med. Phys.16, 627–631~1989!.

111A. S. Meigooni, J. A. Meli, and R. Nath, ‘‘Interseed effects on dose125I brachytherapy implants,’’ Med. Phys.19, 385–390~1992!.

112R. G. Dale, ‘‘Some theoretical derivations relating to the tissue dosimof brachytherapy nuclides’ with particular reference to iodine-125Med. Phys.10, 176–183~1983!.

113L. Mandell, D. Nori, L. Anderson, M. Belanich, and B. Hilaris, ‘‘Theeffects of permanent I-125 interstitial implantation on cortical bone’’ IJ. Radiat. Oncol., Biol., Phys.11, 1029–1034~1985!.

114C. C. Ling and E. D. Yorke, ‘‘Interface dosimetry for I-125 sourcesMed. Phys.16, 376–381~1989!.

115A. S. Meigooni and R. Nath, ‘‘Tissue inhomogeneity correction f

sy

e

g,e

tio

siniol

g,nt

a-t. J.

lys.

ithf-ol.,

25ss.

. J.n-dip-


brachytherapy sources in a heterogeneous phantom with cylindricalmetry,’’ Med. Phys.19, 401–407~1992!.

116D. Y. C. Huang, M. C. Schell, K. A. Weaver, and C. C. Ling, ‘‘Dosdistribution of125I sources in different tissues,’’ Med. Phys.17, 826–832~1990!.

117E. D. Yorke, D. Y. C. Huang, M. C. Schell, R. Wong, and C. C. Lin‘‘Clinical implication of I-125 dosimetry of bone and bone-soft tissuinterfaces,’’ Int. J. Radiat. Oncol., Biol., Phys.21, 1613–1619~1991!.

118R. G. Dale, ‘‘The application of the linear-quadratic dose-effect equato fractionated and protracted radiotherapy,’’ Br. J. Radiol.58, 515–528~1985!.

119R. G. Dale, ‘‘Radiobiological assessment of permanent implants utumor repopulation factors in the linear-quadratic model,’’ Br. J. Rad62, 241–244~1989!.

120J. N. Roy, L. L. Anderson, K. E. Wallner, Z. Fuks, and C. C. Lin‘‘Tumor control probability for brachytherapy: Application to permaneprostate implants,’’ Radiother. Oncol.28, 72–75~1993!.

121C. G. Orton, ‘‘Time-dose factors~TDF’s! in brachytherapy.’’ Br. J. Ra-diol. 47, 603–607~1974!.


m-

n

g.

122C. C. Ling, ‘‘Permanent implants using Au-198, Pd-103 and I-125: Rdiobiological considerations based on the linear quadratic model.’’ InRadiat. Oncol., Biol., Phys.23, 81–87~1992!.

123C. C. Ling, W. X. Li, and L. L. Anderson, ‘‘The relative biologicaeffectiveness of I-125 and Pd-103,’’ Int. J. Radiat. Oncol., Biol., Ph32, 373–378~1995!.

124M. L. Freeman, P. Goldhagen, E. Sierra, and E. J. Hall, ‘‘Studies wencapsulated125I sources. II. Determination of the relative biological efectiveness using cultured mammalian cells,’’ Int. J. Radiat. OncBiol., Phys.8, 1355–1361~1982!.

125M. J. Marchese, E. J. Hall, and B. S. Hilaris, ‘‘Encapsulated iodine-1in radiation oncology. I. Study of the relative biological effectivene~RBE! using low dose rate irradiation of mammalian cell cultures,’’ AmJ. Clin. Oncol.7, 607–611~1984!.

126M. J. Marchese, P. E. Goldhagen, M. Zaider, D. J. Brenner, and EHall, ‘‘The relative biological effectiveness of photon radiation from ecapsulated iodine-125, assessed in cells of human origin: I. Normalloid fibroblasts,’’ Int. J. Radiat. Oncol., Biol., Phys.18, 1407–1413~1990!.

permanent prostate seed implant brachytherapy: report of ... · association of physicists in...

Documents