Comparison of seed loading approaches in prostate brachytherapy Wayne M. Butlera) Schiffler Oncology Center, Wheeling Hospital, 1 Medical Park, Wheeling, West Virginia 26003-6300
Gregory S. Merrick Schiffler Oncology Center, Wheeling Hospital, 1 Medical Park, Wheeling, West Virginia 26003-6300 and The George Washington University Medical Center, Division of Radiation Oncology & Biophysics, 901 23rd Street NW, Washington, DC 20037-2377
Jonathan H. Lief Schiffler Oncology Center, Wheeling Hospital, 1 Medical Park, Wheeling, West Virginia 26003-6300 and Wheeling Jesuit University, 316 Washington Avenue, Wheeling, West Virginia 26003-6295
Anthony T. Dorsey Schiffler Oncology Center, Wheeling Hospital, 1 Medical Park, Wheeling, West Virginia 26003-6300
共Received 14 January 1999; accepted for publication 24 November 1999兲 Since uniform seed loading in prostate brachytherapy can produce an intolerably high dose along the urethra, some form of peripheral loading is commonly employed. We define three variants of peripheral loading and compare them in a small, medium, and large prostate in terms of coverage of the planning target volume 共PTV兲, homogeneity, and ability to spare critical structures of excessive dose. Modified uniform loading has at least 2/3 of the seeds occupying sites on a 1 cm cubic grid keyed to the prostate base and the posterior border of the prostate. Nonuniform loading explicitly spares the urethra by using only basal and apical seeds in at least two centrally located needles. Peripheral loading uses higher activity seeds with the posterior implant plane 5 mm anterior to the posterior border of the prostate. The three prostate volumes 共18.7, 40.7, and 60.2 cm3 by ultrasound兲 were expanded to planning volumes 共32.9, 60.0, and 87.8 cm3, respectively兲. The planning volumes 共PTVs兲 were loaded with a 125I seed distribution and activity sufficient to cover 99.7⫾0.3% of the PTV with the prescribed minimal peripheral dose 共mPD兲 of 145 Gy. Activities used ranged from 0.32 to 0.37 mCi/seed 共0.41–0.47 U/seed兲 for the first two approaches and from 0.57 to 0.66 mCi 共0.72–0.84 U兲 for peripheral loading. Modified uniform loading produced the most uniform distribution based on dose–volume histograms and the volume receiving ⬎150% of prescribed dose. All the approaches are capable of constraining the superior–inferior dose profile 共the urethral path兲 to less than 150% of the mPD, but the nonuniform approach with explicit urethral sparing kept the urethral dose below 120% of the mPD. Dose profiles for the three approaches along the posterior–anterior midline axis are comparable near the urethra, but peripheral and nonuniform approaches have extended regions where the dose is ⬎150% of mPD. These regions approach within 10 mm of the rectum or urethra, so these two approaches require greater accuracy in intraoperative execution of the plan. Although each of the three planning approaches can achieve the treatment goals of adequate coverage and critical structure sparing, modified uniform loading has a more homogeneous dose distribution. This approach may be more forgiving of systematic errors in seed placement. © 2000 American Association of Physicists in Medicine. 关S0094-2405共00兲01202-5兴 Key words: prostate brachytherapy, iodine-125, dosimetry, DVH I. INTRODUCTION Over the past decade, transperineal ultrasound guided prostate brachytherapy has emerged as a viable option in the management of early stage carcinoma of the prostate gland. Measured in terms of local control and freedom from biochemical failure, the results of conformal brachytherapy have been found to be as favorable as the most positive radical prostatectomy series.1–6 Initially, the Seattle group utilized a uniform seed loading philosophy.1,4,7 However, because a purely uniform loading scheme produces a high central dose which may adversely affect the urethra,8 this loading philosophy has evolved into an approach utilizing fewer central seeds and more peripheral seeds. A task group of the Ameri381
Med. Phys. 27 „2…, February 2000
can Association of Physicists in Medicine 共AAPM兲 recommends that treatment plans be ‘‘designed to place seeds peripherally to improve dose homogeneity and to avoid unnecessary radiation damage to the urethra.’’ 9 Currently, most brachytherapy programs utilize a spectrum of loading approaches which satisfy the AAPM criteria and which may be classified according to our definition in Sec. II as either modified uniform loading,10,11 peripheral loading,3,12 or nonuniform loading.13 All three approaches are reasonable, and none has been proven clinically superior in terms of either clinical outcomes or side effects/ complications. Herein, we present a theoretical comparison of the three approaches with emphasis on the potential advantages and disadvantages of each.
0094-2405Õ2000Õ27„2…Õ381Õ12Õ$17.00
© 2000 Am. Assoc. Phys. Med.
381
382
Butler et al.: Comparison of seed loading approaches
II. METHODS AND MATERIALS A. Ultrasound volumeÕplanning volume
Transverse images were recorded every 0.5 cm from the proximal seminal vesicles/base to the apex. The base, or 0.0 cm transverse image plane, was taken to be where the base of the seminal vesicles are clearly visible along with the tip of the base of the prostate capsule. Since the prostate apex does not often occur cleanly at a 0.5 cm increment from the base, the apical slice was usually designated as the next slice caudal to the last slice in which the prostate was clearly visible. Only when the last slice with clearly visible prostate had prostate dimensions less than 1 cm was this extra slice not appended to the study set. Using these criteria for locating the base and apex, even the smallest prostates were planned as 3.5 cm long and the largest implantable glands were planned as 6 cm long. For both clinical and physical reasons, the ultrasound volume was expanded by about 150% to a planning volume 共PTV兲. Pathologically, the incidence of organ confined disease ranges from 64% in patients with prostate specific antigen 共PSA兲 levels between 0 and 4 ng/mL to only 18% in patients with PSA between 20.1 and 30 ng/mL.14 The extracapsular extension often occurs along the neurovascular pedicles at the base and apex. An adequate periprostatic dose, therefore, requires generous enlargement at the base and apex. Physically, needle placement uncertainty is about 5 mm longitudinally and 3 mm in transverse directions.15 This placement uncertainty and the likelihood of extracapsular extension led us to define the cross-sectional area of the PTV on any slice to be at least as large as the area of the ultrasound prostate on the next larger adjoining slice. The PTVs were drawn with bilateral symmetry about the sagittal midline on each image to compensate for small differences in alignment between the volume study and the operative procedure and for prostate movements inherent to the implant procedure and which occur despite stabilization of the gland. To ensure rectal sparing, at midgland there was no expansion of the posterior border which in each case was flat and aligned with the lowest row implantable with the available template. The expansion laterally and anteriorly near midgland was ⭐3 mm unless symmetry considerations dictated otherwise. Even small enlargements add up to large volume increases in inverse proportion to the size of the gland. Total enlargement for the small prostate chosen for this study was 176% (18.7→32.9 cm3), the medium prostate 147% (40.7 →60.0 cm3), and the large prostate 146% (60.2 →87.8 cm3). A prescribed dose of 145 Gy to the PTV helps ensure that the prostate with margin receives 145 Gy as the minimal peripheral dose 共mPD兲.16 B. Modified uniform loading
In our classification scheme, this approach uses seed activities less than 0.4 mCi 共0.51 U兲 and has more than 2/3 of the seeds occupying lattice points on a 1 cm cubic grid with two planes of the lattice defined by the base plane of the Medical Physics, Vol. 27, No. 2, February 2000
382
prostate and the posterior border/lowest implant row. In our practice, the 125I seed activities used for modified uniform loading fall in three groups based on the size of the prostate: small glands 共⬍25 cm3兲 use 0.35⫾0.02 mCi共0.44⫾0.03 U兲, medium glands 共25–45 cm3兲 use 0.34⫾0.02 mCi共0.43 ⫾0.03 U兲, and large glands 共⬎45 cm3兲 use 0.33 ⫾0.02 mCi共0.42⫾0.03 U兲. We have found that the modified uniform loading approach may be implemented by following a simple manual algorithm which consistently covers ⬎98% of the PTV with the prescribed mPD and ⬍35% of the PTV receives 150% of the mPD. This latter measure of homogeneity is commonly referred to as V 150 and the former measure of coverage as V 100 . 16
1. Loading algorithm The algorithm for modified uniform seed loading consists of the placement of needles and seeds on the main uniform grid, followed by modifications which place additional needles and seeds on the lateral and anterior periphery and also reduce the linear density of seeds in central needles. This improves homogeneity and spares the urethra. Starting from the largest transverse image near midgland, the posterior plane of needles aligns with the posterior border of the prostate. A needle may occupy either the intersection of a sagittal 共vertical兲 midline with this plane or a position 0.5 cm lateral to the intersection. This choice identifies the main uniform loading grid as either ‘‘uppercase’’ or ‘‘lowercase’’ when the sagittal midline column on the template is labeled ‘‘D’’ while the columns 0.5 cm lateral to it are labeled ‘‘c’’ and ‘‘d.’’ The plans in this study used lowercase columns for the main grid so that central orthogonal dose profiles would not pass through any seeds. Needles will pass through all valid lattice points on a 1 cm2 grid at midgland that are within the PTV plus any valid grid points which are within a 2 mm margin of the plan area drawn on the midgland slice. These needles extend to the base plane which defines the starting plane for valid seed positions in 10 mm increments along each needle. All needles deposit seeds at the base plane unless the seed would be more than 1 cm outside the plan area at the base. For planes at 1 cm increments from the base, seeds occupy all valid uniform grid points which are no more than 4 mm outside the plan area drawn on that plane except for the apical plane for which a 1 cm margin is allowed. If the uniformly spaced seeds do not fall on the apical plane, the seed trains are extended 0.5 cm inferior to the apex with the plan area of the apical plane used as a guide for valid grid points occupied by seeds. For example, the medium prostate used in this study was 4.5 cm long and required six seeds in needles extending from 0 to 5 cm.
2. Addition of peripheral needles The first modification to uniform loading places needles on the lateral and anterior periphery. Referring once again to the largest transverse image, additional needles are placed along the lateral and anterior periphery on lattice positions that are offset 0.5 cm in both the x and y directions from the uniform loading grid and which are no more than 5 mm
383
Butler et al.: Comparison of seed loading approaches
383
FIG. 1. Representative adjoining pairs of transverse views of the three prostate sizes with seeds loaded according to the modified uniform approach. The two thicker lines are the ultrasound prostate volume and the larger planning volume, and the two thinner lines are the 100% and 150% isodose lines. Top pair: small prostate, 35 and 40 mm 共apex兲 offsets from the base. Middle pair: medium prostate, 20 and 25 mm offsets from the base. Bottom pair: large prostate, 15 and 20 mm offsets from the base.
inside the plan area or 3 mm outside on the midgland image. For example, when the main uniform grid has needles loaded on lowercase columns and half integer rows, the modification will place needles on uppercase columns and whole integer rows around the lateral and anterior periphery. These needles will deposit seeds with a 1 cm spacing, but the starting points will be offset from the base by 0.5, 1.5, or 2.5 cm. The offset is determined by whether the needle is no more than 5 mm inside or outside the plan area on the relevant half integer offset transverse image. At this time all needle paths, both uniform and peripheral, are checked to ensure that they contain at least two seeds with a uniform 1 cm spacing throughout the length of each needle.
3. Removal of central seeds The second modification reduces the linear seed density in the two to four needles closest to the urethra. Small glands with five-seed needles have the middle seed removed 共and Medical Physics, Vol. 27, No. 2, February 2000
replaced with a spacer if using preloaded needles兲 in two central needles. Medium glands have two to four central needles with six seeds converted to five seed needles by replacing the third and fourth seeds with spacers and changing the spacer between those positions to a seed. Large glands which require seven-seed needles have four central needles converted to five seeds by replacing the third and fifth seed with a space or spacers. The obvious intent of this approach with its low seed activity and allowance of seeds outside the PTV is to ‘‘pull’’ dose out to the margins of the volume rather than constraining all seeds to be within the volume and to ‘‘push’’ dose out to the margin by raising seed activity. Seeds are infrequently lost to embolization by this approach and embolization is particularly rare when the peripheral and periprostatic seeds are contained within Rapid Strand™ braided suture material.17,18 Although the algorithm presented above consistently produces good implants in terms of V 100 and V 150 ,
384
Butler et al.: Comparison of seed loading approaches
384
FIG. 2. Coronal views of the medium prostate showing the complete needle loading in the modified uniform loading approach. The most posterior plane is at 1.5 cm and the most anterior at 5.0 cm.
further adjustments may result in slight additional improvement. For the examples in this study, the plans were optimized for all approaches to produce a V 100 of 99.7⫾0.3% based on calculated dose–volume histograms 共DVHs兲. Representative views of a small, medium, and large prostate with modified uniform seed loading are shown in Fig. 1 and the complete needle loading for the medium prostate in Fig. 2. C. Nonuniform loading with explicit urethral sparing
In this approach, less than 2/3 of the seeds are on the uniform grid established by the posterior plane and the base plane as discussed above in modified uniform loading. Unfortunately, this approach is less amenable to detailed algorithmic solution since more than a few seeds must be added Medical Physics, Vol. 27, No. 2, February 2000
or deleted in response to calculations of coverage and homogeneity. Typical urethral sparing is achieved—for example, starting from a modified uniform ‘‘lowercase’’ plan, where the main grid has needles offset 0.5 cm lateral to the vertical midline of the gland—by removing two central needles on small glands or four central needles circumferential to the urethra otherwise. Removal of these central needles creates a central volume with dose less than the mPD which must be counteracted by placement of additional needles on a plane 0.5 cm anterior to the posterior border plane so that needles are 0.5 cm lateral and 0.5 cm anterior to needles in the posterior plane. The tips of these new needles extend to the transverse plane 0.5 cm inferior to the base plane or planes at
385
Butler et al.: Comparison of seed loading approaches
385
FIG. 3. Transverse views of nonuniform loading with explicit urethral sparing for the medium size prostate: 142 seeds at 0.32 mCi 125I, 32 needles. The two thicker lines are the ultrasound prostate volume and the larger planning volume, and the two thinner lines are the 100% and 150% isodose lines. The apex is at 4.5 cm offset, but a capping plane is at 5.0 cm.
half integer centimeter offsets from the base plane. Valid seed positions in these uniformly loaded needles should be no more than 3 mm outside the plan volume. An additional similar horizontal plane of needles should be placed 0.5 cm anterior to the most anterior central needle removed. Two additional vertical needle planes are often established 0.5 cm lateral to the left and right of the removed central needles. These two vertical and two horizontal planes along with the still extant needles from the modified uniform plan create double or triple layers of needles about the holMedical Physics, Vol. 27, No. 2, February 2000
low central core. Where the central needles have been removed, cold spots usually persist at the base and apex. To achieve virtually complete coverage while maintaining low individual seed activity in this study, single seeds were placed at the base and apex using the previously removed needles. Such fine tuning in capping the base and apex is probably not necessary in clinical practice and the accurate placement of individual seeds is technically difficult in the operative procedure secondary to seed placement uncertainties.15
386
Butler et al.: Comparison of seed loading approaches
386
FIG. 4. Transverse views of peripheral loading of the medium prostate: 70 seeds at 0.63 mCi 125I, 13 needles. The two thicker lines are the ultrasound prostate volume and the larger planning volume, and the two thinner lines are the 100% and 150% isodose lines. The apex is at 4.5 cm offset from the base, but a capping plane is at 5.0 cm.
An example of nonuniform loading with explicit urethral sparing is shown in Fig. 3 for the medium size prostate. D. Peripheral loading
As with nonuniform loading, this approach is not readily amenable to algorithmic design but is characterized by keeping almost all seeds inside the PTV and using higher seed strength to push the dose out to the margins. Needles and seeds are also kept at least 1 cm from the urethra. As with the other two approaches, seeds are 1 cm apart along each needle path, but since the peripheral approach requires about twice Medical Physics, Vol. 27, No. 2, February 2000
the individual source strength of the other two approaches, needles are rarely placed closer than 1 cm apart. Furthermore, to spare the rectum, the posterior plane should be 5 mm anterior to the posterior border of the prostate. With the posterior plane of needles established on a midgland slice, an arc of needles should be placed just inside the planning area following the lateral and anterior periphery with each needle at least 1 cm from its nearest neighbor. As with the other approaches, these needles should have seeds extending symmetrically from the base to the apex 共or 0.5 cm inferior to the apex when the apex falls on a half integer
387
Butler et al.: Comparison of seed loading approaches
387
TABLE I. Implant data summary by implant approach and size.
Approach Modified uniform Nonuniform, urethra sparing Peripheral
Number of seeds
Number of needles
Modified uniform Nonuniform, urethra sparing Peripheral
Total activity mCi 共U兲
V 100 共vol %兲
V 150 共vol %兲
99.7 99.9
18.2 27.6
99.8
38.4
99.8 100
17.7 59.7
46.2 (58.7)
99.5
41.5
Large: 60.2 cm3 ultrasound, 87.8 cm3 planning volume 165 29 0.32 (0.41) 52.8 (67.1) 173 33 0.32 (0.41) 55.4 (70.4)
99.8 99.7
22 43.5
99.9
54.6
Small: 18.7 cm3 ultrasound, 32.9 cm3 planning volume 79 17 0.37 (0.47) 29.2 (37.1) 82 21 0.37 (0.47) 30.3 (38.5) 47
10 3
Modified uniform Nonuniform, urethra sparing Peripheral
Activity per seed mCi 共U兲
0.63 (0.80)
29.6 (37.3)
3
Medium: 40.7 cm ultrasound, 60.0 cm planning volume 133 25 0.32 (0.41) 41.2 (52.3) 142 32 0.32 (0.41) 44.0 (55.9) 70
103
13
19
plane兲. In implementing this approach, seeds were still placed outside the PTV to minimize the number of needless and avoid using single seed needles, but the allowable margins are decreased due to the higher activity. At the base and apex, seeds were allowed up to 7 mm outside the PTV but no more than 3 mm outside on other slices. For large glands, several interior needles must be placed to build up the dose centrally. The tips of these needles should extend only to the transverse plane 0.5 cm inferior to the base plane so that the seeds dropped by these needles occupy the half integer planes. A representative peripheral loading of the medium prostate is shown in Fig. 4. III. RESULTS Table I summarizes the implant data by implant approach and size. Figure 5 displays the calculated dose–volume histograms 共DVHs兲 for the three studied approaches applied to the different size prostates. By design, each plan was manually optimized to provide essentially the same coverage as measured by V 100 共the fraction of the PTV covered by the mPD兲 to be 99.7⫾0.3% of the PTV. The maximum slope of the DVH curve is a measure of the homogeneity of the implant, and the modified uniform loading approach is seen on inspection to be consistently the most homogeneous. Even though the DVH curve for nonuniform urethral sparing starts out below that for modified uniform loading for small and large volumes, the greater total activity required for the nonuniform approach makes its slope at the inflection point shallower and the integral dose under the curve beyond the mPD greater. The peripheral loading approach tends to require greater total activity than the other approaches, and its curve always lies above the modified uniform curve. V 150 , the fraction of the prostate volume receiving 150% of the prescribed mPD, is often considered a surrogate measure of homogeneity and also provides a number which may be relevant in interpreting dose-related complications. HowMedical Physics, Vol. 27, No. 2, February 2000
0.66 (0.84)
0.57 (0.72)
58.7 (74.5)
ever, such complications will clearly be dependent on the location of the high dose regions, and for these preplans, sparing of the urethra and rectum are inherent to the implant design. The modified uniform loading approach results in the lowest V 150 for the various sized prostates. Nevertheless, the differences between approaches are probably not of clinical significance since V 150 is close to or less than 50% of the PTV in all cases except peripheral loading of the large prostate and nonuniform loading of the medium prostate. While DVHs of the PTV provide a global overview of the implant, evaluation of the local environment around critical structures is also of value. In actual implants, this evaluation may be done by DVHs or dose–surface histograms 共DSHs兲 or point dose profiles of the urethra and rectum. Due to the curving path of the urethra, small variations in seed loading can result in profound differences in the quality parameters of actual implants regardless of the implant approach. For comparing implant approaches, a uniform straight line following the midline template needle path 1.5 cm anterior to the posterior border of the prostate was taken as an idealized urethra for purposes of calculating dose profiles. Here the posterior border of the prostate was set at row 1.5, so the urethra profile line followed template position D 3.0. All implant needles used in the trial preplans were at least 0.5 cm from this idealized urethra. Postimplant dosimetry has shown that locating a pseudourethra at the geometric center of each transverse CT prostate image gives an average urethral dose profile that is within 10% of the actual dose in modified uniform loading19 and peripheral loading.20 Dose profiles along this linear urethra are illustrated in Fig. 6 for the three prostates and implant approaches studied. These profiles show that the idealized urethra receives at least 100% and less than 140% of the mPD over the entire length of the prostate in all but one instance. Peripheral loading of the medium prostate has seeds capping the base and apex near the profile line which causes the dose to rise to 160% of mPD. Without the artifact of seeds capping the
388
Butler et al.: Comparison of seed loading approaches
388
FIG. 5. Dose–volume histograms for small, medium, and large prostates using the three seed loading approaches. Modified uniform⫽solid line, nonuniform with explicit urethral sparing ⫽dashed line, peripheral⫽dotted line.
ends, the dose would fall to 85% of mPD at the base and apex, concomitantly reducing the V 100 coverage. Another set of profiles orthogonal to the idealized urethra at midgland in the posterior–anterior direction is illustrated in Fig. 7. Although all the loading approaches achieve an adequately low dose near the urethra, approaches which place a larger fraction of the total implanted seed strength near the periphery, the explicit urethral sparing of nonuniform and peripheral approaches, deliver relatively high dose near the anterior and posterior margins. Although the position of the rectum is not explicitly marked in the figure, the anterior rectal wall may lie as close as 3–5 mm from the posterior prostate margin at midgland. Medical Physics, Vol. 27, No. 2, February 2000
IV. DISCUSSION Task Group 56 has recommended that implants not use uniform seed loading but rather tend toward peripheral loading in order to spare the urethra.9 Purely uniform and purely peripheral loading anchor two ends of the dose spectrum of implants delivering the mPD to the target volume between which most clinically planned implants should lie. At one end of the spectrum, the dose is high centrally and falls to the mPD on the periphery, while at the other extreme, an ideal high dose surface lies within the PTV so that the dose falls to the mPD at the periphery and the center. The implant approaches presented here all avoid high central doses, but
389
Butler et al.: Comparison of seed loading approaches
389
FIG. 6. Superior–inferior dose profiles for small, medium, and large prostates using the three seed loading approaches. Modified uniform⫽solid line, nonuniform with explicit urethral sparing⫽dashed line, peripheral ⫽dotted line. The dotted vertical lines indicate the location of the base and apex.
those we label peripheral only crudely approximate the ideal surface of pure peripheral loading. Although each loading approach has a multitude of solutions producing similar results and the implant designs presented here represent only a single sample of each solution set for a single isotope, the samples are broadly representative of the intent of each approach. Modified uniform seed placement is intended to produce a relatively uniform dose throughout the PTV by placing extra seeds on the periphery and removing some central seeds. Nonuniform loading with explicit urethral sparing requires an annulus of higher dose around the central seed exclusion zone in order to bring the dose at the urethra up to Medical Physics, Vol. 27, No. 2, February 2000
the prescribed mPD. Peripheral loading reduces the thickness of this annulus by increasing the individual seed strength. All three seed loading approaches 共modified uniform, nonuniform urethral sparing, and peripheral兲 can accomplish adequate coverage of the treatment PTV with acceptable homogeneity and adequate urethral and rectal sparing. The modified uniform loading approach, however, results in greater dose homogeneity measured in terms of DVH slope and orthogonal dose profiles. There is no clinical data regarding the efficacy of more homogeneous permanent brachytherapy implants or the prevention of undesired treatment sequelae, but calculations of tumor control probability
390
Butler et al.: Comparison of seed loading approaches
390
FIG. 7. Posterior–anterior dose profiles for small, medium, and large prostates using the three seed loading approaches. Modified uniform⫽solid line, nonuniform with explicit urethral sparing⫽dashed line, peripheral ⫽dotted line. The dotted vertical lines indicate the location of the posterior and anterior margin of the planning volume.
in prostate brachytherapy21 indicate that in 125I implants with D99 共the minimal dose covering 99% of the prostate volume兲 of 100 Gy, doses higher than 130 Gy result in no additional cell kill. All three approaches produce comparable sparing of the urethra and rectal wall. We noted in Fig. 7 that the peripheral approach utilizing high activity seeds and the nonuniform approach using closely spaced layers of seeds near the periphery produce high dose regions in close proximity to critical structures. As such, precise seed placement is essential for these two approaches while modified uniform loading is the least dependent on accurate seed placement to spare critical structures. Utilizing a modified uniform approach with relatively low Medical Physics, Vol. 27, No. 2, February 2000
activity seeds makes accurate seed placement less crucial, and misplacement of a single seed results in little change in the dose distribution as shown by Roy et al.8 in comparing uniform loading to peripheral loading. The modified uniform approach is also less sensitive to other relatively common errors which involve mistargeting single needles or systematically misaligning the target volume by a few millimeters. Due to the low activity seeds used in modified uniform and nonuniform loading, errors in the placement of single needles are unlikely to have significant adverse consequences, but systematic errors in alignment of a nonuniform or peripheral plan may place a high dose region at the urethra or rectum. Furthermore, misplacement of a single high activ-
Butler et al.: Comparison of seed loading approaches
391
ity peripherally located seed may have greater dosimetric consequences. Wallner, Roy, and Harrison12 utilized a peripheral loading approach with relatively high activity seeds, and they reported that doses to the urethra in excess of 400 Gy increase the risk of urethral morbidity. In addition, not only the dose but the length of the urethra irradiated was important in Wallner’s series. Patients who developed Radiation Therapy Oncology Group 共RTOG兲 grade 0-1 toxicity received 400 Gy to ⭐10 mm of the urethral length whereas patients with RTOG grade 3–4 toxicity received on average 400 Gy to a urethral length of 20 mm. This suggests that misplaced high activity sources may significantly increase the dose delivered to the urethra and result in increased morbidity. To achieve such a high dose over such an extended distance with the lower activity seeds used in the other approaches requires multiple seeds within 5 mm of each other and adjacent to the urethra. Although there are no isodose volumes contiguous between seed positions at 200% of mPD in any of the examples shown here, such contiguous high isodose volumes are common in nonuniform loading implants due to intentional close spacing of needles in an annulus, but contiguous 200% mPD isodose volumes occur rarely in modified uniform implants. In peripheral and nonuniform loading, systematic mistargeting of the gland will not only not benefit the patient, but may lead to harm as high dose regions overlie the urethra. The only clear clinical advantage to peripheral loading is in patients who have previously undergone transurethral resections of the prostate gland 共TURP兲. Ragde et al.4 reported a 31% incidence of urinary incontinence in patients who have previously undergone a TURP. Since this patient population was treated via a uniform seed loading approach with subsequent high central doses, alternative implant approaches sparing the epithelium of the TURP defect have been implemented. A recent publication by Wallner et al.22 reported only a 6% chance of urinary incontinence at a median follow-up of 36 months in TURP patients implanted with relatively high activity seeds in a peripheral loading technique with exquisite attention to urethral sparing. With regards to adequate coverage, Stock et al.23 have recently reported a dose response curve for 125I monotherapy. They reported optimal biochemical control for 125I monotherapy patients who received a D 90 共minimal dose covering 90% of the prostate volume兲 of 140–160 Gy. Their implant technique is similar to the nonuniform urethral sparing approach with a contiguous annular high dose region. Over a four-year period Stock et al. gradually increased the total activity implanted per unit volume in order to consistently achieve D 90’s greater than 140 Gy. Increasing the source activity will always improve a suboptimal coverage parameter but the improvement is accompanied by greater V 150 . Theoretically, all three approaches result in acceptable dose distributions, but the modified uniform approach is ‘‘more forgiving’’ of local and systematic errors in seed placement due to greater dose homogeneity within the implanted volume. However, this approach, along with nonuni-
form loading, uses a larger number of needles and low activity seeds which increases the cost of the procedure in terms of time and material compared to higher activity peripheral loading. On the other hand, the use of a peripheral approach with high activity seeds or a nonuniform approach may result in significant volumes of overdosage and the possibility of the placement of high dose regions over critical structures subsequent to misplacement of seeds.
391
Medical Physics, Vol. 27, No. 2, February 2000
a兲
Electronic mail:
[email protected] J. C. Blasko, P. D. Grimm, and H. Ragde, ‘‘Brachytherapy and organ preservation in the management of carcinoma of the prostate,’’ Sem. Rad. Oncol. 3, 240–249 共1993兲. 2 K. Wallner, J. Roy, and L. Harrison, ‘‘Tumor control and morbidity following transperineal I-125 implantation for stage T1/T2 prostatic carcinoma,’’ J. Clin. Oncol. 14, 449–453 共1996兲. 3 M. Dattoli et al., ‘‘Pd-103 brachytherapy and external beam radiation therapy for clinically localized high risk prostatic carcinoma,’’ Int. J. Radiat. Oncol., Biol., Phys. 35, 875–879 共1996兲. 4 H. Ragde, J. C. Blasko, P. D. Grimm, G. M. Kenny, J. E. Sylvester, D. C. Hoak, K. Landin, and W. Cavanagh, ‘‘Interstitial I-125 radiation without adjuvant therapy in the treatment of clinically localized prostate cancer,’’ Cancer 共N.Y.兲 80, 442–453 共1997兲. 5 H. Ragde, J. C. Blasko, P. D. Grimm, G. M. Kenny, J. Sylvester, D. C. Hoak, W. Cavanagh, and K. Landin, ‘‘Brachytherapy for clinically localized prostate cancer: Results at 7 and 8 year follow-up,’’ Sem. Surg. Oncol. 13, 438–443 共1997兲. 6 H. Ragde, A. A. Elgamal, P. B. Snow, J. Brandt, A. A. Bartolucci, B. S. Nadir, and L. J. Korb, ‘‘Ten year disease free survival after transperineal sonography-guided iodine-125 brachytherapy with or without 45-gray external beam irradiation in the treatment of patients with clinically localized, low to high Gleason grade prostate carcinoma,’’ Cancer 共N.Y.兲 83, 989–1001 共1998兲. 7 P. D. Grimm, J. C. Blasko, and H. Ragde, ‘‘Ultrasound guided transperineal implantation of iodine-125 and palladium-103 for the treatment of early stage prostate cancer,’’ Atlas. Urol. Clin. N. Am. 2, 113–125 共1994兲. 8 J. N. Roy, C. C. Ling, K. E. Wallner, and L. L. Anderson, ‘‘Determining source strength and source distribution for a transperineal prostate implant,’’ Endocurietherapy/Hyperthermia Oncol. 12, 35–42 共1996兲. 9 R. Nath, L. L. Anderson, J. A. Meli, A. J. Olch, J. A. Stitt, and J. F. Williamson, ‘‘Code of practice for brachytherapy physics: Report of the AAPM Radiation Therapy Committee Task Group No. 56,’’ Med. Phys. 24, 1557–1598 共1997兲. 10 G. S. Merrick, W. M. Butler, A. T. Dorsey, and H. L. Walbert, ‘‘Prostatic conformal brachytherapy: I-125/Pd-103 post-operative dosimetric analysis,’’ Radiat. Oncol. Invest. 5, 305–313 共1997兲. 11 G. S. Merrick, W. M. Butler, A. T. Dorsey, and H. L. Walbert, ‘‘The influence of timing on the dosimetric analysis of transperineal ultrasound guided prostatic conformal brachytherapy,’’ Radiat. Oncol. Invest. 6, 182–190 共1998兲. 12 K. Wallner, J. Roy, and L. Harrison, ‘‘Dosimetry guidelines to minimize urethral and rectal morbidity following transperineal I-125 prostate brachytherapy,’’ Int. J. Radiat. Oncol., Biol., Phys. 32, 465–471 共1995兲. 13 N. N. Stone and R. G. Stock, ‘‘Brachytherapy for prostate cancer: Realtime three-dimensional interactive seed implantation,’’ Tech. Urol. 1, 72–80 共1995兲. 14 A. W. Partin, M. W. Kattan, E. N. Subong, P. C. Walsh, K. J. Wojno, J. E. Oesterling, P. T. Scardino, and J. D. Pearson, ‘‘Combination of prostate-specific antigen, clinical stage, and Gleason score to predict pathological stage of localized prostate cancer. A multi-institutional update,’’ J. Am. Med. Assoc. 277, 1445–1451 共1997兲. 15 P. L. Roberson, V. Narayana, D. L. McShan, R. J. Winfield, and P. W. McLaughlin, ‘‘Source placement error for permanent implant of the prostate,’’ Med. Phys. 24, 251–257 共1997兲. 16 G. S. Merrick, W. M. Butler, A. T. Dorsey, and J. H. Lief, ‘‘The potential role of various dosimetric quality indicators in prostate brachytherapy,’’ Int. J. Radiat. Oncol., Biol., Phys. 44, 717–724 共1999兲. 17 E. M. Tapen, J. C. Blasko, P. D. Grimm, H. R. Ragde, R. Luse, S. Clifford, J. Sylvester, and T. W. Griffin, ‘‘Reduction of radioactive seed 1
392
Butler et al.: Comparison of seed loading approaches
embolization to the lung following prostate brachytherapy,’’ Int. J. Radiat. Oncol., Biol., Phys. 42, 1063–1067 共1998兲. 18 G. S. Merrick, W. M. Butler, A. T. Dorsey, J. H. Lief, and M. L. Benson, ‘‘Seed fixity in the prostate/periprostatic region following brachytherapy,’’ Int. J. Radiat. Oncol., Biol., Phys. 共in press兲. 19 G. S. Merrick, W. M. Butler, A. T. Dorsey, and H. L. Walbert, ‘‘Prostatic conformal brachytherapy: I-125/Pd-103 post-operative dosimetric analysis,’’ Radiat. Oncol. Invest. 5, 305–313 共1997兲. 20 F. Waterman and A. Dicker, ‘‘Determination of the urethral dose in prostate brachytherapy when the urethra cannot be visualized in the postimplant CT scan,’’ Med. Phys. 共in press兲.
Medical Physics, Vol. 27, No. 2, February 2000
392 21
C. C. Ling, J. Roy, N. Sahoo, K. Wallner, and L. Anderson, ‘‘Quantifying the effect of dose inhomogeneity in brachytherapy: Application to permanent prostatic implant with 125I seeds,’’ Int. J. Radiat. Oncol., Biol., Phys. 28, 971–978 共1994兲. 22 K. Wallner, H. Lee, S. Wasserman, and M. Dattoli, ‘‘Low risk of urinary incontinence following prostate brachytherapy in patients with a prior transurethral prostate resection,’’ Int. J. Radiat. Oncol., Biol., Phys. 37, 565–569 共1997兲. 23 R. G. Stock, N. N. Stone, A. Tabert, C. Iannuzzi, and J. K. DeWyngaert, ‘‘A dose response study for I-125 prostate implants,’’ Int. J. Radiat. Oncol., Biol., Phys. 41, 101–108 共1998兲.
