Int. J. Radiation Oncology Biol. Phys., Vol. 52, No. 2, pp. 544 –552, 2002 Copyright © 2002 Elsevier Science Inc. Printed in the USA. All rights reserved 0360-3016/02/$–see front matter
PII S0360-3016(01)02672-4
PHYSICS CONTRIBUTION
PROSTATE BRACHYTHERAPY POSTIMPLANT DOSIMETRY: A COMPARISON OF PROSTATE QUADRANTS SABEENA SIDHU, B.SC., W. JAMES MORRIS, M.D., INGRID SPADINGER, PH.D., MIRA KEYES, M.D., MICHELLE HILTS, M.SC., ROBERT HARRISON, M.SC., KARL OTTO, M.SC., MICHAEL MCKENZIE, M.D., AND ALEXANDER AGRANOVICH, M.D. BC Cancer Agency, Vancouver, British Columbia, Canada Purpose: To investigate postimplant dosimetry for different regions of the prostate gland in patients treated with transperineal 125Iodine brachytherapy implants for low- and intermediate-risk prostate cancer. Methods and Materials: Two hundred eighty-four patients treated with permanent interstitial prostate brachytherapy comprised the study population. A nonuniform, urethral-sparing algorithm was used to plan all patients. Prostate contours were outlined on postimplant CT images. Prostate volumes were then divided into four quadrants: anterior-superior quadrant (ASQ), posterior-superior quadrant (PSQ), anterior-inferior quadrant (AIQ), and posterior-inferior quadrant (PIQ). Dose–volume histograms (DVHs) were calculated for the whole prostate and each quadrant. Results: The mean postimplant V100 ⴞ 95% confidence (the percent prostate volume encompassed within the isodose surface comprising the prescription dose ⴝ 144 Gy) for the ASQ was 78.5 ⴞ 1.9, which was significantly lower than that of the PSQ, AIQ, and PIQ in which the V100 ⴞ 95% confidence values were 94.9 ⴞ 0.8, 92.6 ⴞ 1.2, and 98.7 ⴞ 0.3, respectively. The mean V100 ⴞ 95% confidence for the whole prostate was 90.4 ⴞ 0.8. Mean values for V150 and D90 (the minimum dose in Gy received by 90% of the target volume) for the four quadrants and the whole prostate showed similar results. Conclusions: Underdosed areas of the planning target volume (PTV), if present, were largely confined to the ASQ, which received a significantly lower dose, on average, compared to the other three quadrants of the prostate. © 2002 Elsevier Science Inc. Prostate brachytherapy, Postimplant dosimetry, Treatment volume.
volume (PTV), qualitative observation of the postimplant dosimetry in our patients suggested that areas of underdose were not randomly distributed, but confined largely to the anterior base of the gland. In this study, we quantify the observed nonrandom distribution of underdosed regions using results obtained from computed tomography (CT)-based postimplant dosimetry. Dose parameters (V100, V150—the percent volume of the prostate that receives ⱖ 144 Gy [the prescribed dose] and ⱖ 216 Gy, respectively, and D90—the minimum dose received by 90% of the prostate volume) are reported for the whole prostate as well as the ASQ, PSQ, AIQ, and PIQ quadrants.
INTRODUCTION The use of permanent transperineal iodine-125 and palladium-103 brachytherapy for prostate cancer offers patients with early-stage disease a curative treatment that is convenient, cost-effective, and less invasive than surgery. Acute urinary morbidity, including outlet obstruction and nocturia, is the most frequent and troublesome side effect following prostate brachytherapy (1, 2). Evidence suggests that urinary side effects are correlated to the size of the prostate and the dose received by the urethra (1–3). Uniform loading of the radioactive sources results in a very high central, and therefore urethral, dose (4). For this reason many brachytherapy centers, including this one, use some form of peripheral loading of the sources to reduce the dose to the urethra (5). However, sparing the central region from excess dose involves some increased risk of underdosing parts of the prostate gland. Even though pretreatment ultrasound planning ensured 99% (143 Gy or more) dose coverage to the planning target
METHODS AND MATERIALS Two hundred ninety-five patients underwent prostate brachytherapy at the Vancouver Cancer Center (VCC) from July 1998 to August 2000. Eleven patients were excluded from this analysis because postimplant CT scans were not
Reprint requests to: W. James Morris, BC Cancer Agency, 600 West 10th Avenue, Vancouver, British Columbia, Canada, V5Z 4E6. Tel: (604) 877-6000 (2673); Fax: (604) 877-0505; E-mail:
[email protected]
This work was supported in part by an unrestricted educational grant from Nycomed Amersham. Received Dec 21, 2000, and in revised form Jul 10, 2001. Accepted for publication Sep 12, 2001. 544
Postimplant quadrant dosimetry
Table 1. Pretreatment clinical characteristics for the 284 study patients* Clinical characteristics PSA score* ⬍ 1 ng/mL ⱖ 1–4 ng/mL ⱖ 4–10 ng/mL ⬎ 10 ng/mL Gleason 4 Gleason 5 Gleason 6 Gleason 7 Hormone treatment Yes No
Number (%) 8 (3) 55 (19) 177 (63) 44 (15) 16 (6) 41 (14) 181 (64) 46 (16) 213 (75) 71 (25)
* PSA scores are those taken before hormone therapy.
available (10 patients) or because bilateral hip prostheses made accurate identification of the prostate contour impossible (1 patient). Patients’ age ranged from 49 to 80 with a mean of 65.5 years. All patients were treated according to a protocol with strict eligibility criteria. No patients received external beam radiation or had supplementary implants to compensate for errors in dose coverage. However, 75% of the patients in this study received neoadjuvant and adjuvant androgen suppression. In all cases where androgen suppression was used, the planning ultrasound study was done after a period of volume reduction lasting at least 3 months. Sealed 125I sources (IMC6711, Nycomed Amersham, Arlington Heights, IL) with an apparent activity of 0.33 mCi (NIST 99) were used for all procedures. A Multimedia Medical Systems (MMS) VariSeed 6.7 (Build 812) package was used for treatment planning and for CT-based postimplant dosimetry analysis (Varian Medical Systems, Charlottesville, VA). Patient eligibility Patients with low-risk prostate cancer (prostate-specific antigen [PSA] ⱕ 10, Gleason score ⱕ 3 ⫹ 3 ⫽ 6, International Union Against Cancer [UICC] 1997 clinical stages T1c–T2b) were eligible for implantation. Those with intermediate-risk features (PSA ⬍ 10 and Gleason score 3 ⫹ 4 ⫽ 7, or PSA 10 –15 and Gleason score 6 or less) were eligible with at least 3 months neoadjuvant androgen suppression and 3 months adjuvant androgen suppression. Patients with high-risk prostate cancer (PSA ⬎ 15, or PSA ⱕ 10 and Gleason score ⱖ 4 ⫹ 3 ⫽ 7) were not considered for brachytherapy. Use of androgen suppression in low-risk patients did not exclude them from brachytherapy. All patients with an implant volume ⬎ 50 cm3 at preoperative assessment received hormone therapy to reduce gland size. Table 1 lists the pretreatment clinical characteristics of the 284 study patients. Treatment planning A typical treatment plan is shown in Fig. 1. A Siemens Endo-PII Transducer (5/7.5 MHz) transrectal probe was
● S. SIDHU et al.
545
used to obtain axial images of the prostate 0.5 cm apart. The images were captured on videotape and downloaded into the VariSeed program in which a radiation oncologist outlined the PTV. The PTV included the entire prostate and extended 0.5–1.0 cm superiorly to include the base of the seminal vesicles and the perivesical soft tissues. Inferiorly, a margin of 0.5 cm was added to the prostate apex. This ensured 100% volume coverage of the entire prostate. Volumes were not extended laterally except to ensure a symmetric contour with respect to the midsagittal plane of the prostate. The VariSeed system was used for all calculations, treatment planning, therapy visualization, and documentation. Treatment planning was performed manually using an algorithm developed at the VCC. As seen in Fig. 1, the planning algorithm involves a nonuniform source distribution. The planned prescription dose of 144 Gy (as per TG 43) typically extended 3–5 mm beyond the PTV in the lateral and posterior regions of the prostate. While all sections of the PTV were planned to receive ⱖ99% of the prescribed dose, a particular effort was made to exclude the urethra and periurethral tissues from doses greater than 216 Gy. This resulted in a relatively small planned V150 within the anterior part of the PTV compared to the posterior part. Preimplant dosimetry Preimplant planning ultrasound PTVs on 30 of the 284 patients were selected and subjected to a quadrant by quadrant analysis of V100, V150, and D90 to fully characterize the nonuniform dose distribution inherent in the VCC planning algorithm. Selection was done by first placing all patients into groups of 10 according to implant date. Then, one patient was randomly selected from each group. For each ultrasound contour in the sample population, the prostate was divided into the four quadrants (ASQ, PSQ, AIQ, and PIQ) using a method identical to that described below for the postimplant CT prostate volumes (see “Postimplant dosimetry”). We are confident that this sample is representative of the whole group, because the planning criteria varied little during the time these patients were planned and treated. Implant procedure The implant procedure was performed on an outpatient basis under general or spinal anesthesia in the dorsal lithotomy position using 18-gauge needles preloaded with 125I sources and spacers. The sources were implanted transperineally through predetermined template apertures using realtime transrectal ultrasound guidance, supplemented by fluoroscopy on demand. In the operating room, left and right needles were placed alternately, and fine adjustments in needle placement and the position of the ultrasound probe were used throughout the procedure. This minimized errors in source placement due to longitudinal and rotational motions of the prostate (6), as well as swelling of the gland during the procedure (7). In many cases, 2– 6 additional sources were placed in regions of known positive biopsies or potential underdose.
546
I. J. Radiation Oncology
● Biology ● Physics
Volume 52, Number 2, 2002
Fig. 1. A typical VCC ultrasound treatment plan showing the modified peripheral loading used in this series of implants. Prostate contours are outlined and shaded. The isodose lines are 144 Gy (prescription dose) and 216 Gy. The preimplant prostate volume was 46.75 cm3. The plan utilized 108 125I seeds having a source strength of 0.33 mCi/seed and 28 needles.
Postimplant quadrant dosimetry
Table 2. Preimplant dosimetry mean values for the prostate and four quadrants of the prostate with their estimated 95% confidence intervals (CI) for the sample of 30 patients treated with brachytherapy
V100 (%) 95% CI V150 (%) 95% CI D90 (Gy) 95% CI Volume (cc) 95% CI
ASQ
PSQ
AIQ
PIQ
Whole
99.2 0.5 50.9 4.3 177.8 3.9 8.5 1.0
99.8 0.2 74.7 5.5 198.4 5.1 8.2 0.9
99.7 0.2 47.8 4.7 178.6 3.5 8.4 0.9
99.9 0.2 70.8 5.6 196.2 5.4 7.9 0.9
99.6 0.2 60.8 3.9 187.5 3.5 33.0 3.2
Postimplant dosimetry Axial plane CT images 3 mm thick were obtained of the pelvic area at 2.5-mm intervals approximately 30 days (mean ⫽ 30, range: 12–70) after implantation. This time period was chosen to allow the acute swelling of the gland to subside (7). The prostate and adjacent organs were outlined on each CT slice by one of two radiation oncologists. The 125I sources were identified on the CT images using a combination of manual selection and automated redundancy checks available on the VariSeed system. After the prostate contours were identified, the prostate volume was divided into four quadrants: ASQ, PSQ, AIQ, and PIQ. This was accomplished using a simple two-step method. First, the CT prostate volume was divided into superior and inferior segments by determining the axial midplane of the prostate. In cases where an uneven number of axial images contained prostate contours, the extra contour was assigned to the inferior quadrants. The anterior and posterior regions were then created on each prostate contour by determining the midpoint along the anterior-posterior axis with a flexible ruler applied directly to the image on the computer screen. The V100, V150, and D90 were recorded for each of the quadrants and the whole prostate. RESULTS Table 2 summarizes the planned dosimetry parameters (V100, V150, and D90) for the entire PTV and the four quadrants as determined from the sample of 30 patients described in “Methods and Materials”. Table 2 shows that all sampled plans specified that the entire PTV and all four quadrants received ⬎99% of the prescription dose. However, from the inception of our program, the specified V150 of the posterior quadrants was planned at high levels (mean planned V150 for PSQ ⫽ 74.7% and for the PIQ ⫽ 70.8%). At the same time, the V150 to the anterior quadrants was planned at relatively low levels (mean planned V150 for ASQ ⫽ 50.9% and for the AIQ ⫽ 47.8%) in an effort to spare the urethra/bladder neck from doses ⬎150% of the prescription dose. The postimplant CT dosimetry results (V100, V150, and
● S. SIDHU et al.
547
Table 3. Postimplant dosimetry mean values for the prostate and four quadrants of the prostate with their estimated 95% CI for the sample of 284 patients treated with brachytherapy
V100 (%) 95% CI V150 (%) 95% CI D90 (Gy) 95% CI Volume (cc) 95% CI
ASQ
PSQ
AIQ
PIQ
Whole
78.5 1.9 31.9 1.7 130.6 2.9 10.7 0.3
94.9 0.8 66.1 2.2 173.5 3.4 10.7 0.3
92.6 1.2 47.7 2.3 161.7 3.0 7.9 0.2
98.7 0.3 76.7 2.0 198.6 3.5 7.8 0.3
90.4 0.8 54.5 1.6 163.6 2.2 37.1 1.1
D90) for the whole prostate and the four defined quadrants are summarized in Table 3. On average, the full prescription dose was received by ⬎90% of the postimplant CT volume for the whole prostate and all quadrants (whole prostate mean V100 ⫽ 90.4%, PSQ ⫽ 94.9%, AIQ ⫽ 92.6%, PIQ ⫽ 98.7%) except the ASQ where the mean V100 was 78.5%. Concordant results were found for the other dosimetry parameters. For example, the mean V150 for the whole prostate was 54.5%, whereas the mean V150 of the ASQ was only 31.9%. The mean V150 values of the other three quadrants were appreciably higher (PSQ ⫽ 66.1%, AIQ ⫽ 47.7%, and PIQ ⫽ 76.7%). Similarly, the mean D90 of the whole gland was 163.6 Gy, while that of the PSQ ⫽ 173.5 Gy, the AIQ ⫽ 161.7 Gy, the PIQ ⫽ 198.6 Gy, and the ASQ ⫽ 130.6 Gy. Thus, for each measured parameter, the dose to the ASQ was significantly lower than that of the other quadrants and the whole prostate. Figures 2 through 4 are frequency histograms of the three measured postimplant CT dose parameters for all 284 patients comprising the study population. For V100, the frequency plots show that the measured values are tightly grouped above 90% for the PSQ, PIQ, and AIQ, whereas there is a much broader range of V100 values for the ASQ. However, the narrow range of values seen for the V100 for the PSQ, PIQ, and AIQ is largely an artifact, because, by definition, the V100 cannot exceed 100%. The V150 suffers less from this limitation since the V150 very rarely exceeded 100% and the frequency plots for V150 (Fig. 3) have a broader range of values. Furthermore, the range is similar in all four quadrants and the whole prostate despite marked differences in their mean values. The minimum dose received by 90% of the postimplant volume (D90, see Fig. 4) has no upper or lower limit and is, therefore, the best single value by which to estimate the range of variance seen in our study results. As in the case of V150, the D90 values show a similar variance for all four quadrants and the whole prostate, despite significant differences in their respective mean values. Forty-six patients (16%) had a whole prostate D90, which was less than the prescription dose of 144 Gy, and 199 patients (70%) had a D90 value ⬍ 144 Gy for the ASQ. Potters et al. (8) identified that the D90 dose ⱖ 90% of the prescribed dose can be used as a factor for predicting PSA
548
I. J. Radiation Oncology
● Biology ● Physics
Volume 52, Number 2, 2002
Fig. 2. Histogram distributions of the V100 for the (a) ASQ, (b) PSQ, (c) AIQ, (d) PIQ, and (e) whole prostate. Data were sorted into bins of 1%. Not shown in (a): 5 patients with a V100 ⬍ 40%.
relapse-free survival in patients treated with brachytherapy. In this study sample, 13 patients (4.6%) had a D90 value for the whole prostate of ⱕ130 Gy. This was markedly different for the ASQ, where 139 patients (49%) had a D90 value ⱕ 130 Gy. At the other extreme, 48 patients (17%) had D90 values ⱖ 180 Gy for the whole prostate and 9 patients (3.2%) had D90 values ⱖ 200 Gy. Such high doses were uncommon in the ASQ for which only 8 patients (2.8%) had D90 values ⱖ 180 Gy.
DISCUSSION Our results demonstrate that the anatomic locations of underdose (i.e., areas of the CT prostate volume receiving less than the prescribed dose) are not randomly or uniformly distributed. The ASQ is more likely than the other quadrants to contain areas of relatively low dose. There are several reasons that may account for the low dose in this particular quadrant, and these are discussed below.
Postimplant quadrant dosimetry
● S. SIDHU et al.
549
Fig. 3. Histogram distributions of the V150 for the (a) ASQ, (b) PSQ, (c) AIQ, (d) PIQ, and (e) whole prostate. Data were sorted into bins of 1%.
First, the planning algorithm employed at the VCC deliberately specifies a high source density in the posterior peripheral regions of the prostate. At the same time, a relatively low source density is planned for the anterior half of the prostate to avoid excessive dose to the urethra and bladder neck (2, 9, 10). Small errors in source placement in the anterior quadrants will, therefore, have a proportionally greater impact on the resulting postimplant CT dosimetry than placement errors of similar magnitude in the posterior quadrants.
Second, needle drag on the prostate while pulling back and depositing sources can result in sources being placed inferiorly to the intended coordinates. During implant procedures, we make every effort to compensate for this commonly recognized cause of source misplacement. Nevertheless, needle drag may have contributed to the relatively low doses delivered to the ASQ. Indirect evidence for this effect can be seen in Tables 2 and 3, where the postimplant dosimetric parameters for the AIQ fall very close to the preimplant specifications despite the fact that both the AIQ
550
I. J. Radiation Oncology
● Biology ● Physics
Volume 52, Number 2, 2002
Fig. 4. Histogram distributions of the D90 for the (a) ASQ, (b) PSQ, (c) AIQ, (d) PIQ, and (e) whole prostate. Data were sorted into bins of 1 Gy.
and ASQ were planned with similar source densities as discussed above. This implies that some of the sources intended for the ASQ may have been deposited inferiorly to the intended coordinates and ended up in the AIQ through the effects of needle drag. Similarly, the dose received by the PIQ (mean D90 ⫽ 198.6 Gy) is higher than the PSQ (mean D90 ⫽ 173.5 Gy) despite equal and high planned source densities for both quadrants, once again suggesting that needle drag may have resulted in an inferior shift of some sources. Third, the lower dose in the ASQ could also be due to
needle splay, which can have serious dosimetric effects. Nath et al. (11) found that the dosimetric quality of an implant degraded with needle divergence. For 125I implants, the average reduction in minimum target dose was about 10% and 20% for a needle divergence of 5° and 10°, respectively. Furthermore, peripheral needles may be subject to severe splaying, which mainly affects the distribution of sources superiorly. A study by Roberson et al. (12) found that the average placement error of sources dropped in the base of the prostate was 6 mm. Peripheral sources are more
Postimplant quadrant dosimetry
difficult to position precisely and are often deposited in places where high dose fall-off occurs. This lowers the quality of the implant, which may result in a loss of tumor control. During the procedure at our institution, adjustments were made in real time to compensate for shifts in the relative positions of the prostate, needle tip, and the intended X-Y coordinates to minimize this effect. Nevertheless, needle splay is more pronounced in the needles intended for anterolateral coordinates and, therefore, could result in a greater risk of source misplacement and consequent underdose in the ASQ compared to other regions. Difficulty in accurately contouring the prostate on postimplant CT scan is well-recognized (13, 14). This is a particular problem in the ASQ where the anterior-superior CT prostate contour is difficult to separate from the periprostatic venous plexus and blends with the bladder neck. Inadvertent inclusion of these tissues in the CT prostate contour would result in extension of the ASQ volume into regions that were not implanted. The relatively low dosimetric parameters of the ASQ could, therefore, reflect inclusion of this additional volume, rather than underdose of the actual ASQ of the prostate. Finally, source loss through the mechanisms of embolization and loss in the voided urine could contribute to the relatively high frequency of underdosed areas in the ASQ. Embolization occurs principally when sources are inadvertently inserted into the periprostatic venous plexus, which lies in close proximity to the ASQ. In the present study, 30-day postimplant chest and pelvic X-rays were available in 192 of the 284 patients in this study. Sixty-six of these patients (34%) had source emboli in the lungs. The mean number of sources found in the lungs was 2 (range 1–5, median ⫽ 1). In addition, 18 patients (9.4%) had sources identified in the pelvis, but outside the intended implant volume. In total, 17,368 sources were implanted in these 192 patients. Of these, 138 sources (0.8%) were identified on 30-day postimplant X-rays of the chest and pelvis as being outside the intended implant volume. Because of the proximity of the venous plexus to the ASQ, it is probable that the majority of these sources were intended for implant in the ASQ. Source loss in the voided urine is likewise a possible cause of underdose in the ASQ. It is common knowledge that patients occasionally have sources in the bladder at the end of the procedure, and many centers undertake cystoscopy to recover such sources or provide patients with special instructions and kits to strain their urine 1–3 weeks after the implant. At our center we do neither, so we have no direct measure of the number of sources lost in the urine. However, we have estimated the number of sources lost in the voided urine by assuming that all sources that were not identified on the pelvic or chest X-rays were, in fact, passed in the urine. In our sample of 192 patients, 95 patients (49%) had one or more sources that could not be accounted for on the pelvic and chest X-rays. In total, 187 of 17,368 implanted sources (1.1%) were apparently voided in the urine between the date of implant and the 30-day postimplant X-rays. Because of the intimate
● S. SIDHU et al.
551
anatomic relationship between the bladder neck/base and the ASQ, it is reasonable to speculate that most of the sources lost in the voided urine were intended for the ASQ. Adding the potential losses in the voided urine with those from embolization, we have calculated that 1.9% (325/ 17,368) of all the implanted sources were lost from the intended implant volume. Because 58 patients experienced no loss of sources, the remaining 134 patients experienced, on average, a loss of 2.6% of the intended sources (325 out of 12,477 sources implanted in these 134 patients). If we are correct in our speculation that these losses came predominantly from sources intended for the ASQ, then source loss (combined venous embolization and losses in the urine) could be a contributing factor to the underdose of the ASQ. We have quantified a well-known observation in prostate brachytherapy, namely that the anterior-superior part of the gland is more likely than other regions to contain areas where the dose delivered is less than the prescribed dose. However, the clinical significance of underdosed areas in the ASQ is yet to be determined. There have been several histologic studies of the prostate (14 –17) that suggest a minority of adenocarcinomas originate in the ASQ and that spread of cancer to this region is unlikely. D’Amico et al. (15) recommend excluding the anterior base from receiving 100% of the prescription dose particularly for low-risk patients with a prostatic volume less than 35 cm3. A lower dose to the ASQ would significantly lower urethral dose, which may reduce the probability of urethral toxicity and morbidity. Supporting this view is a separate study by Bucci et al. (3) of the same 284 patients reported in this paper. These patients were analyzed for their acute urinary morbidity (first 6 months postimplant) as scored by the patientreported International Prostate Symptom Score (IPSS) and the physician-reported Radiation Therapy Oncology Group (RTOG) grade. In this study, multivariate analysis showed that only the dose to the ASQ, the preimplant IPSS, and the ratio of the postimplant CT volume to the ultrasound planning target volume (a surrogate for postimplant swelling and edema) were significant predictors of acute urinary morbidity. However, further clinical investigation is required to determine if intentionally reducing the dose to the ASQ will affect tumor control. Although some data have demonstrated an increased risk of treatment failure in patients receiving relatively low doses to the prostate as a whole (8, 18), there have been no studies that specifically correlate dose to the ASQ with biochemical recurrence or other measures of clinical outcome. Also, because the ASQ is already getting less than the prescribed dose in many patients in our series, intentionally underdosing the ASQ may have undesirable consequences in neighboring prostate quadrants. In our treatment planning, we continue to ensure ⱖ99% dose coverage of all four prostate quadrants, and do not feel it would be prudent to intentionally underdose the ASQ in the absence of clinical data demonstrating equal efficacy and reduced toxicity. Further investigations of toxicity and tumor control for the patients reported in this study
552
I. J. Radiation Oncology
● Biology ● Physics
are currently under way. Once adequate follow-up is available for this group, we hope to be able to assess the clinical consequences of the relative underdose identified in the
Volume 52, Number 2, 2002
ASQ. It may then be possible to make a more informed recommendation regarding a strategy of deliberate dose reduction to the ASQ.
REFERENCES 1. Kleinberg L, Wallner K, Roy J, et al. Treatment-related symptoms during the first year following transperineal 125I prostate implantation. Int J Radiat Oncol Biol Phys 1994;28:985–990. 2. Wallner K, Roy J, Harrison L. Dosimetry guidelines to minimize urethral and rectal morbidity following transperineal I-125 prostate brachytherapy. Int J Radiat Oncol Biol Phys 1995;32:465– 471. 3. Bucci J, Morris WJ, Keyes M, et al. Predictive factors of acute urinary symptoms following prostate brachytherapy (abstract). J Brachyther. In press. 4. Narayana V, Roberson PL, Winfield RJ, et al. Optimal placement of radioisotopes for permanent prostate implants. Radiology 1996;199:457– 460. 5. Bice WS, Prestidge BR, Grimm PD, et al. Centralized multiinstitutional post-implant analysis for interstitial prostate brachytherapy. Int J Radiat Oncol Biol Phys 1998;41:921–927. 6. Blasko JC, Grimm PD, Ragde H. Brachytherapy and organ preservation in the management of carcinoma of the prostate. Sem Rad Onc 1993;3:240 –249. 7. Prestidge BR, Bice WS, Kiefer EJ, et al. Timing of computed tomography– based post-implant assessment following permanent transperineal prostate brachytherapy. Int J Radiat Oncol Biol Phys 1998;40:1111–1115. 8. 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;50:605– 614. 9. Butler WM, Merrick GS, Lief JH, et al. Comparison of seed loading approaches in prostate brachytherapy. Med Phys 2000;27:381–392.
10. Nath R, Anderson LL, Meli JA, et al. Code of practice for brachytherapy physics: Report of the AAPM Radiation Therapy Committee Task Group No. 56. Med Phys 1997;24:1557– 1598. 11. Nath S, Chen Z, Yue N, et al. Dosimetric effects of needle divergence in prostate seed implant using 125I and 103Pd radioactive seeds. Med Phys 2000;27:1058 –1066. 12. Roberson PL, Narayana V, McShan DL, et al. Source placement error for permanent implant of the prostate. Med Phys 1997;24:251–257. 13. Sandler HM, Bree RL, McLaughlin PW, et al. Localization of the prostatic apex for radiation therapy using implanted markers. Int J Radiat Oncol Biol Phys 1993;27:915–919. 14. McNeal JE. Normal histology of the prostate. Am J Surg Pathol 1988;12:619 – 633. 15. D’Amico AV, Davis A, Vargas SO, et al. Defining the implant treatment volume for patients with low risk prostate cancer: Does the anterior base need to be treated? Int J Radiat Oncol Biol Phys 1999;43:587–590. 16. McNeal JE, Redwine EA, Freiha FS, et al. Zonal distribution of prostatic adenocarcinoma: Correlation with histologic pattern and direction of spread. Am J Surg Pathol 1988;12:897– 906. 17. McNeal JE, Price H, Redwine EA, et al. Stage A versus stage B adenocarcinoma of the prostate morphologic comparison and biologic significance. J Urol 1988;139:61– 65. 18. Stock RG, Stone NN, Tabert A, et al. A dose-response study for I-125 prostate implants. Int J Radiat Oncol Biol Phys 1998;41:101–108.
