Brachytherapy 6 (2007) 129e134

A prospective randomized comparison of stranded vs. loose125I seeds for prostate brachytherapy

Daniel R. Reed1,2,3,*, Kent E. Wallner2,3,4, Gregory S. Merrick5, Sandra Arthurs3,Amy Mueller3, William Cavanagh3, Wayner B. Butler5, Eric Ford2,3, Steve G. Sutlief2,3

1Radiation Oncology, Arizona Oncology Services, Phoenix, AZ2Department of Radiation Oncology, University of Washington, Seattle, WA

3Radiation Oncology, Puget Sound Health Care System, Department of Veterans Affairs, Seattle, WA4Radiation Oncology, Group Health Cooperative, Seattle, WA

5Schiffler Cancer Center, Wheeling, WV

ABSTRACT PURPOSE: To compare seed loss and dosimetric parameters between stranded and loose 125Iseeds (LS) for prostate brachytherapy.METHODS AND MATERIALS: Sixty-four patients with 1997 American Joint Commission onCancer (AJCC) clinical stage T1c or T2a prostate carcinoma were prospectively randomized to bra-chytherapy (144 Gy) with RAPID Strand 125I seeds (RS) vs. LS (Oncura, Plymouth Meeting, PA)The treatment plan for each patient was devised before randomization, and was not modified basedon the randomization. Each patient underwent magnetic resonance, computed tomography, andplain film radiographs on the day of the implant (Day 0) and 30 days later (Day 30).RESULTS: Overall, 21 of 62 patients (30%) experienced seed loss. Seed loss occurred in 15 of 32of LS patients (47%) vs. 6 of 30 RS patients (23%; p 5 0.053). Mean seed loss was 1.09 in the LSpatient vs. 0.43 in RS patients ( p 5 0.062). Eight LS patients (25%) lost multiple seeds, comparedto 3 stranded patients (10%). Despite the lesser degree of seed loss in patients who receivedstranded seeds, they had a paradoxical trend toward lower V100 and D90 values.CONCLUSION: This prospective randomized trial showed a strong trend toward a decrease inpostimplant seed loss with stranded seeds. Improved seed retention may be more advantageousin a setting of less generous periprostatic coverage. The lowered risk seed migration seen withstranded seeds would presumably also decrease the likelihood of lung or cardiac seed embolization.� 2007 American Brachytherapy Society. All rights reserved.

Keywords: Prostate; Brachytherapy; Dosimetry; Seeds

Introduction

Maximizing the likelihood of biochemical control withprostate brachytherapy is partly dependent on delivery ofan adequate radiation dose to the prostate volume (1e3).Additionally, there is evidence that generous treatment mar-gins (TMs) may maximize the likelihood of cancer eradica-tion, presumably by encompassing carcinoma that haspenetrated the prostate capsule (4). Unfortunately,

Received 28 June 2006; received in revised form 17 January 2007;

accepted 17 January 2007.

Financial support: Oncura, Plymouth Meeting, PA.

* Corresponding author. Arizona Oncology Services, Scottsdale Radi-

ation Oncology, 8994 E. Desert Cove Avenue, Suite 100, Scottsdale, AZ

85260. Tel.: þ1-602-274-4484; fax: þ1-4800-314-3343.

E-mail address: dreed@azoncology.com (D.R. Reed).

1538-4721/07/$ e see front matter � 2007 American Brachytherapy Society

doi:10.1016/j.brachy.2007.01.003

a dosimetrically optimal implant is not always achieved,due to minor source placement errors, edema, and seed mi-gration or seed loss (5e9). Additionally, highly peripheralseed placement, with a large percentage of periprostaticand extra-glandular seeds, may increase the incidence ofsource migration (5e18).

One potential way to improve seed placement and reten-tion is the use of stranded seeds. Suture-mounted 125I sour-ces (RAPID Strand, Oncura, Plymouth Meeting, PA) areheld in #1 Vicryl (polyglactin 910). The suture is stiffenedand sterilized with ethylene oxide gas before being shippedinside a plastic spacing jig, packed inside a stainless steelshielding tube (Fig. 1). Theoretically, suture-mounted sour-ces are more likely to maintain their position, and shouldnot migrate if one or two sources of the suture are protrud-ing partly out of the prostate. They may decrease seed

. All rights reserved.

130 D.R. Reed et al. / Brachytherapy 6 (2007) 129e134

Fig. 1. Suture-mounted 125I sources (RAPID Strand) are shipped inside a plastic spacing jig, which is shipped inside a stainless steel shielding tube. The

strand is cut between the sources with a razor-sharp blade, and then picked up with tweezers, grasping only the portion that holds a source (31).

movement within and outside the prostate, presumably en-hancing prostatic and periprostatic coverage.

In the authors’ experience, stranded seeds result in whatseems qualitatively to be superior maintenance of seed po-sition over time (Fig. 2). However, multiple retrospective

comparisons of stranded vs. loose sources have shown con-flicting conclusions regarding their dosimetric benefits(19e22). Accordingly, we conducted a prospective ran-domized trial to compare dosimetric outcomes with loosevs. stranded sources.

Fig. 2. Plain en face view of implants with loose seeds vs. stranded seeds at Day 0 (top panels) and Day 30 (lower panels). Seed spacing and alignment seem

to be better maintained with the stranded product.

131D.R. Reed et al. / Brachytherapy 6 (2007) 129e134

Methods and materials

Patient characteristics

Sixty-four patients with 1997 AJCC clinical stage T1c orT2a prostate carcinoma were prospectively randomized tobrachytherapy with RAPID Strand 125I seeds (RS) vs. loose125I seeds (LS) at the Puget Sound VA from March 2003through May 2004. The patients were treated and consentedin accordance with University of Washington Human Sub-jects policy. Randomization was assigned by a computer-ized random number generator to one of the twotreatment arms.

The prescription dose was 144 Gy (TG-43). One patientfrom the stranded seed arm did not return for followupfilms and 1 patient randomized to LS was never treated,leaving 62 patients for analysis (Table 1). Treatment planswere designed with a modified peripheral loading pattern,with approximately one third of seeds being placed extra-capsularly (5, 23).

LS were hand-inserted with standard techniques usinga Mick applicator (17). RS were inserted through standardMick needles (without the Mick applicator) under fluoro-scopic guidance. The treatment plan for each patient wasdevised before randomization, and was not modified basedon the randomization. All anterior, lateral, and posteri-orelateral needles contained alternating seeds, amenable

Table 1

Patients excluded from current analysis

Loose seeds RAPID Strand Total

Total randomized 33 31 64

Never treated 1 1

No return visit 1 2

Patients analyzed 32 30 61

to strands. All plans included two posterior medial rowsof five or six seeds at irregular spacing that were loadedwith LS from a Mick applicator. Additionally, strandedplans typically contained 10e20 single, LS primarily atthe base and apical region. In summary, RS plans includeda total of 16e38 LS (median, 26) in addition to the strandedseeds, with LS accounting for 25e45% of the total(median, 33%; Fig. 3).

Each patient underwent magnetic resonance (MR), com-puted tomography (CT), and plain film radiographs on theday of the implant (Day 0) and 30 days later (Day 30). TheVariSeed 7.1 software package (Varian Medical Systems,Inc., Palo Alto, CA) was used to import and fuse MR im-ages to the CT data set on which dosimetry calculationswere performed. The fusion algorithm consists of a rigidbody transformation based on user-defined points in bothimage sets. For points, we used bony landmarks (5e10per patient) located as close as possible to the prostate.

Table 2

Treatment variables

Loose seeds RAPID Strand p Value

Patient number 32 30

Prostate volume (cc) 32 (�12) 33 (�9) 0.98

mCi/seed 0.52 (�0.018) 0.51 (�0.016) 0.49

Total activity (mCi) 39 (�8.5) 39 (�6.3) 0.85

Day 0 seed number 75.2 (�17) 74.8 (�12) 0.93

Table 3

Seed loss summary

Loose seeds RAPID Strand p Value

Mean seeds lost 1.1 (�1.6) 0.4 (�1.0) 0.062

Patients with seed loss 15/32 (47%) 7/30 (23%) !0.053

Fig. 3. Typical plan used for loose or RAPID Strand patients. For stranded plans, the irregularly spaced seeds (red) were placed with a Mick applicator. This

patient had a total of 76 seeds, 20 of which were loose and 56 stranded.

132 D.R. Reed et al. / Brachytherapy 6 (2007) 129e134

The resulting fusion was reviewed in the interactive displaywindow and minor manual adjustments were made wherenecessary. Dose calculations were based on the CT images,with interpretative guidance from the MR images. Seeddefinition was accomplished on the CT images, becauseseeds are not clearly visible on MR images. Image fusionfor all cases was performed by one physicist (EF), and allanatomic delineation was performed by one physician(KEW). A manual seed identification and redundancycheck was performed for all cases.

Dosimetric parameters analyzed included the V100 andV200, defined as the percent of the postimplant prostate vol-ume covered by 100% and 200% of the prescription dose,respectively. The D90 was defined as the minimal dose re-ceived by 90% of the prostate. The rectum was delineatedas a single contour around the outer rectal wall on eachtransverse image containing prostate volume to calculatethe R100dthe volume of rectum receiving 100% or moreof prescribed dose (24). No extra imaging (chest X-rays,etc.) was performed to determine the sites of seedmigration.

The study was originally designed to enroll 74 patients,giving an 80% power to detect a 2.5-mm difference in TM

Fig. 4. Seed loss from Day 0 to Day 30 for loose vs. stranded seeds.

achieved. Enrollment was terminated at 64 patients, when itbecame apparent that TMs were not a practical endpoint.

Results

Patient and treatment variables including, preimplantprostate volume, isotope strength (mCi), Day 0 seed num-ber, and total activity were closely matched between groups(Table 2). Overall, 21 of 62 patients (30%) experiencedseed loss. Seed loss occurred in 15 of 32 of LS patients(47%) vs. 6 of 30 RS patients (23%; p 5 0.053) (Table3). Mean seed loss was 1.09 in the LS patient vs. 0.43 inRS patients (p 5 0.062). Eight LS patients (25%) lost mul-tiple seeds, compared to 3 stranded patients (10%) (Fig. 4).No strands were lost.

The Day 0 V100 and D90 values were nearly identical be-tween groups. The mean Day 0 V100 was 95% for both LSand RS patients (p 5 0.96). Similarly, the mean Day 0 D90

was 169 Gy for both LS and RS patients (p 5 0.88). De-spite the lesser degree of seed loss in patients who receivedstranded seeds, they had a paradoxical trend toward lowerV100 and D90 values. The mean Day 30 V100 was 96% forLS patients vs. 94% for RS patients (p 5 0.12). The meanDay 30 D90 was 178 Gy for LS vs. 164 Gy for RS(p 5 0.058) (Fig. 5). Overall, there was a trend towardlower V100 and D90 values on Day 30 with increasingnumber of seeds lost (Fig. 6).

The mean R100 on Day 0 and Day 30 was 0.56 and1.96 cc, respectively, for the LS arm vs. 0.81 and 1.91 ccfor the RS arm (Table 4).

Discussion

The results presented here demonstrate a strong trend to-ward fewer seeds lost in patients implanted with strandedvs. loose seeds. Patients treated with RS had a 23% riskof seed loss, compared to a 47% incidence of seed loss in

Fig. 5. Change in V100 (left) and D90 (right) from Day 0 to Day 30.

133D.R. Reed et al. / Brachytherapy 6 (2007) 129e134

Fig. 6. Seeds lost vs. Day 30 V100 (left) and Day 30 D90 (right).

patients implanted with LS ( p 5 0.053). The difference inthe percent of patients losing seeds, however, did not trans-late into a clinically significant difference in postimplantdosimetry. We attribute the lack of dosimetric effects partlyto the low absolute number of seeds lost in either group,and to our use of high numbers of extra-prostatic seedsand very generous TMs (23). Highly peripheral preplanswith a large percentage of extra-prostatic seeds allow forvery generous postimplant TMs, which are robust and resis-tant to decreases in dosimetric parameters despite someseed loss (25).

The paradoxical decrease in Day 30 dosimetric parame-ters in stranded seed patients despite their lesser seed lossmay represent a tendency of stranded seeds to stay periph-eral, decreasing intraprostatic high dose regions. Minor dif-ferences in dosimetric outcomes over time are difficult toexplain, and may stem from complex prostate shapechanges that are not easily quantifiable (26). Similarly, in-creases in rectal doses from Day 0 to Day 30 were similarbetween loose and stranded seeds, and were consistent withprior reports (27).

One caveat to this study is that one third of seeds in thestranded patients were actually LS, to accommodate ourtypical seed placement patterns. It is likely that the seedloss difference would have been greater if less LS wereused in the patients randomized to stranded seeds.

This prospective randomized trial showed a significantdecrease in postimplant seed loss with stranded seeds.The lesser seed loss did not, however, result in improvedDay 30 dosimetry in the setting of highly peripheral seed

Table 4

Change in rectal parameters on Day 0 vs. Day 30

Loose seeds (cc) RAPID Strand p Value

Day 0 R100 0.56 0.81 0.25

Day 30 R100 1.96 1.91 0.8

R100 change 1.4 1.1 0.3

placement with generous TMs (23, 25, 28e30). Improvedseed retention may be more advantageous in a setting ofless generous periprostatic coverage. The lowered risk ofseed migration seen with stranded seeds would presumablyalso decrease the likelihood of lung or cardiac seed embo-lization (15, 16, 18). This study was not designed, however,to quantify the distant sites of seed migration.

Conclusion

There is a trend toward decreased postimplant seed losswith stranded seeds. A more definitive study would includemore patients, less LS included in the stranded plans, fol-lowup chest X-rays to better characterize seed loss, longerfollowup, and fiducial markers to facilitate imageregistration.

References

[1] Potters L, Cao Y, Calugaru E, et al. A comprehensive review of CT-

based dosimetry parameters and biochemical control in patients

treated with permanent prostate brachytherapy. Int J Radiat Oncol

Biol Phys 2001;50:605e614.

[2] Stock RG, Stone NN, Dahlal M, et al. What is the optimal dose for

125-I prostate implant? A dose-response analysis of biochemical

control, posttreatment prostate biopsies, and long-term symptoms.

Brachytherapy 2002;1:83e89.

[3] Wallner K, Merrick G, True L, et al. I-125 versus Pd-103 for low risk

prostate cancer: Preliminary biochemical outcomes from a prospec-

tive randomized multicenter trial. Int J Radiat Oncol Biol Phys

2003;57:1297e1303.

[4] Choi S, Wallner KE, Merrick G, et al. Treatment margins predict bio-

chemical outcomes after prostate brachytherapy. Cancer J 2004;10:

175e180.

[5] Nedea E, Wallner K, Reed D, et al. Extra-prostatic seed placement

and its effect on seed loss. Cancer J 2005;11:147e151.

[6] Al-Qaisieh B, Carey B, Ash D, et al. The use of linked seeds elimi-

nates lung embolization following permanent seed implantation for

prostate cancer. Int J Radiat Oncol Biol Phys 2004;59:1063e1067.

134 D.R. Reed et al. / Brachytherapy 6 (2007) 129e134

[7] Tapen EM, Blasko JC, Grimm PD, et al. Reduction of radioactive

seed embolization to the lung following prostate brachytherapy. Int

J Radiat Oncol Biol Phys 1998;42:1063e1067.

[8] Kunos C, Resnick M, Kinsella T, et al. Migration of implanted free

radioactive seeds for adenocarcinoma of the prostate using a Mick

applicator. Brachytherapy 2004;3:71e77.

[9] Blair H, Porter A, Chen Q-S. In vivo detection of an 125-I seed lo-

cated in the intracardiac region after prostate permanent brachyther-

apy. Int J Radiat Oncol Biol Phys 2004;58:888e891.

[10] Sommerkamp H, Rupprecht M, Wannenmacher M. Seed loss in inter-

stitial radiotherapy of prostatic carcinoma with I-125. Int J Radiat

Oncol Biol Phys 1988;14:389e392.

[11] Stutz M, Petrikas J, Raslowsky M, et al. Seed loss through the

urinary tract after prostate brachytherapy: examining the role of

cystoscopy and urine straining post implant. Med Phys 2003;30:

2695e2698.

[12] Dafoe-Lambie J, Abel L, Blatt H, et al. Radioactive seed emboliza-

tion to the lung following prostate brachytherapy. W V Med J

2000;96:357e360.

[13] Nag S, Vivekanandam S, Martinez-Monge R. Pulmonary emboliza-

tion of permanently implanted radioactive palladium-103 seeds for

carcinoma of the prostate. Int J Radiat Oncol Biol Phys 1997;39:

667e670.

[14] Steinfeld AD, Donahue BR, Plaine L. Pulmonary embolization of io-

dine-125 seeds following prostate implantation. Urology 1991;37:

149e150.

[15] Davis BJ, Pfeifer EA, Wilson TM, et al. Prostate brachytherapy seed

migration to the right ventricle found at autopsy following acute

cardiac dysrhythmia. J Urol 2000;164:1661.

[16] Davis B, Bresnahan J, Stafford S, et al. Prostate brachytherapy seed

migration to a coronary artery found during angiography. J Urol

2002;168:1103.

[17] Ankem M, DeCarvalho V, Harangozo A, et al. Implications of radio-

active seed migration to the lungs after prostate brachytherapy. Urol-

ogy 2002;59:555e559.

[18] Eshleman J, Davis B, Pisansky T, et al. Radioactive seed migration to

the chest after transperineal interstitial prostate brachytherapy: ex-

traprostatic seed placement correlates with migration. Int J Radiat

Oncol Biol Phys 2004;59:419e425.

[19] Kaplan I, Meskell P, Lieberfarb M, et al. A comparison of the preci-

sion of seeds deposited as loose seeds versus suture embedded seeds:

A randomized trial. Brachytherapy 2004;3:7e9.

[20] Fuller D, Koziol J, Feng A. Prostate brachytherapy seed migration

and dosimetry: analysis of stranded sources and other potential pre-

dictive factors. Brachytherapy 2004;2004:10e19.

[21] Fagundes H, Keys R, Wojcik M, et al. Transperineal TRUS-guided

prostate brachytherapy using loose seeds versus RAPIDStrand: A

dosimetric analysis. Brachytherapy 2004;3:136e140.

[22] Lee R, deGuzman AF, Tomlinson S, et al. Radioactive sources em-

bedded in suture are associated with improved post-implant dosime-

try in men treated with prostate brachytherapy. Radiother Oncol

2002;65:123e127.

[23] Merrick GS, Butler WM. Modified uniform seed loading for prostate

brachytherapy: rationale, design, and evaluation. Tech Urol 2000;6:

78e84.

[24] Mueller A, Wallner K, Merrick G, et al. Perirectal seeds as a risk

factor for prostate brachytherapy-related rectal bleeding. Int J Radiat

Oncol Biol Phys 2004;59:1047e1052.

[25] Merrick GS, Butler WM, Dorsey AT, et al. The dependence of pros-

tate postimplant dosimetric quality on CT volume determination. Int

J Radiat Oncol Biol Phys 1999;44:1111e1117.

[26] McLaughlin P, Narayana V, Pan C, et al. Comparison of day 0 and

day 14 dosimetry for permanent prostate implant using stranded

seeds. Int J Radiat Oncol Biol Phys 2005;64:144e150.

[27] Taussky D, Yeung I, Williams T, et al. Rectal-wall dose dependence

on postplan timing after permanent-seed prostate brachytherapy. Int J

Radiat Oncol Biol Phys 2006;65:358e363.

[28] Butzbach D, Waterman FM, Dicker AP. Can extraprostatic extension

be treated by prostate brachytherapy? An analysis based on postim-

plant dosimetry. Int J Radiat Oncol Biol Phys 2001;51:1196e1199.

[29] Merrick GS, Butler WM, Wallner KE, et al. Extracapsular dose

distribution following permanent prostate brachytherapy. Am J Clin

Oncol 2003;26:178e189.

[30] Reed D, Wallner K, Ford E, et al. Effect of post-implant edema on

prostate brachytherapy treatment margins. Int J Radiat Oncol Biol

Phys 2005;63:1469e1473.

[31] Wallner K, Merrick G, Blasko J, et al. Prostate brachytherapy made

complicated. 3rd ed. Seattle: SmartMedicine Press; 2007.