- Email: [email protected]
Evaluation of the visibility of a new thinner 125I radioactive source for permanent prostate brachytherapy
Brachytherapy 11 (2012) 460e467
Evaluation of the visibility of a new thinner 125I radioactive source for permanent prostate brachytherapy Gemma Roberts, Bashar Al-Qaisieh*, Peter Bownes Medical Physics and Engineering, St James’s Institute of Oncology, St James’s Hospital, Leeds, UK
ABSTRACT
PURPOSE: The 125I source currently used for prostate brachytherapy at St. James’s Institute of Oncology is a standard size seed (z4.5 mm in length and 0.8 mm in diameter). A new, thinner seed is under evaluation. This is designed to be implanted using narrower needles, potentially reducing edema and improving the dose distribution. This study investigated the visibility of the thinner source on multimodality images and compared it with that of standard size seeds. METHODS AND MATERIALS: Images of dummy seeds of both thinner and standard size models were taken using ultrasound, fluoroscopy, computed tomography (CT), and magnetic resonance (MR) imaging. The ultrasound, fluoroscopy, and CT images were acquired with the seeds inserted into phantoms positioned in a water tank. The MR images were acquired using phantoms containing single seeds. The images were analyzed visually and quantitatively. The resolution of closely spaced seeds on CT images was investigated. RESULTS: The visibility of both seeds was similar on ultrasound, fluoroscopy, and MR images. On CT images, the thinner seeds give reduced artifacts and better resolution. CONCLUSIONS: The use of the thinner seed would have minimal effect on ultrasound and fluoroscopy imaging during treatment. However on CT images, the use of the thinner seeds may improve seed identification for post-treatment dosimetry. Further study is required into the suitability of MR images alone for post-treatment dosimetry. Ó 2012 American Brachytherapy Society. Published by Elsevier Inc. All rights reserved.
Keywords:
Prostate brachytherapy;
125
I; OncoSeed Model 9011; THIN Seed; Visibility
Background and purpose Early stage prostate cancer may be treated with permanent brachytherapy using 125I sources (1). The seed currently used at St. James’ Institute of Oncology (SJIO) is the OncoSeed model 6711 (2) (Oncura Inc., Arlington Heights, IL). The seeds are supplied sutured into strands of 10 (RAPIDStrand; Oncura Inc.), which can be cut to size. OncoSeed model 6711 is used in many clinical reviews of treatment outcome currently available (3, 4). The OncoSeed model 6711 source Received 23 September 2011; received in revised form 12 January 2012; accepted 12 January 2012. Financial disclosure: The authors above have no conflict of interest in connection with the article and the material described is not under publication or consideration for publication elsewhere. The non-radioactive dummy seeds for this imaging study were provided by Oncura Inc., Arlington Heights, IL). The study was designed and carried out independently of Oncura Inc. * Corresponding author. St James’s Institute of Oncology, St James’s Hospital, Beckett Street, Leeds LS9 7TF, UK. Tel.: þ44-113-206-7409. E-mail address: [email protected] (B. Al-Qaisieh).
consists of an active core of 125I adsorbed onto the surface of a silver rod (length 2.8 mm, diameter 0.508 mm) contained within a titanium capsule (length 4.56 mm, diameter 0.774 mm) (5). The OncoSeed model 6711 seeds are implanted using 18-G needles, which have an outer diameter of 1.27 mm. A new, thinner seed model (OncoSeed model 9011dTHIN Seed; Oncura Inc., Amersham, UK) has recently been developed and our center has obtained some non-radioactive dummy seeds for evaluation. The OncoSeed 9011 model seed is of a similar design to the 6711 and is manufactured in a similar way (6). The OncoSeed model 9011 seed will also be available in sutured strands (THIN Strand; Oncura Inc.). The main difference is the smaller diameter (source 0.30 mm, capsule 0.508 mm) (Fig. 1), which allows narrower 20-G needles (0.91 mm outer diameter) to be used during the implant process. This will potentially reduce the level of edema and trauma to the prostate and urethra (7). The prostate volume may change during and after the implant process because of edema (8). Several studies have
1538-4721/$ - see front matter Ó 2012 American Brachytherapy Society. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.brachy.2012.01.009
G. Roberts et al. / Brachytherapy 11 (2012) 460e467
461
Fig. 1. Cross-sectional diagram showing the dimensions of the OncoSeed 6711 and OncoSeed 9011 (THIN Seed) (Image courtesy of Oncura Inc.).
shown that edema can affect the percentage of the prostate receiving the prescribed dose (V100) by moving the seeds and increasing the volume (9e13). V100 and D90 (the dose received by 90% of the prostate) have been linked to patient outcome (3, 4, 14, 15). Assuming equivalent dosimetry, the new seed model could therefore potentially improve the dose coverage to the prostate by reducing the level of edema. The study of Buskirk et al. (16) suggests that needle trauma is the main cause of short-term urinary symptoms. Several other studies (17e22) have found a statistically significant correlation ( p 5 0.025) (18) between needle trauma during implant (via the number of needles) and edema and urinary retention, which suggests that the new narrower needles could prove advantageous. At the Seattle Prostate Institute, Seattle, WA, model 9011 OncoSeeds have been in clinical use since August 2008 (22). Initial results based on the first 100 patients treated have shown improved results for post-treatment dosimetry when compared with previous patients treated with model 6711 seeds, similar urinary retention rates, and lower levels of pain and bruising. Several imaging modalities are used in the treatment and followup of 125I prostate brachytherapy patients. Transrectal ultrasound (TRUS) is used for guiding needle insertion and in some centers for imaging seeds after implant (23). Fluoroscopy screening can be used at various stages during implant to check the positioning of the seeds. Computed tomography (CT) and magnetic resonance imaging (MRI) scans are used to identify seed positions for post-implant dosimetry. Before bringing a new source into clinical use, it is important to ensure that images obtained with the source are fit for purpose. This study aims to assess whether the introduction of model 9011 seeds and the associated 20-G narrow bore needle will have a clinically significant impact on the appearance and localization of the seeds using any of the above imaging techniques. The importance of source visibility is mentioned by Heintz et al. (23) and Meigooni et al. (24) 2009, but
only a small number of studies characterizing prostate brachytherapy sources take account of this aspect. Previous work at this center by Al-Qaisieh et al. (25) compared the visibility of five seeds implanted in a tissue equivalent phantom and imaged using ultrasound, CT, MRI, and fluoroscopy. Clinically significant differences in visibility were not found in this study. The present study is a continuation of the previous work by Al-Qaisieh et al. and was initiated as a result of the new THIN Seed source coming to market. The two studies jointly represent a comprehensive review of the visibility of 125 I sources from some of the more common manufacturers in Europe (Oncura Inc., Nucletron B.V, Veenendaal, Netherlands and Eckert & Ziegler BEBIG sa, Seneffe, Belgium). This is the first study looking at the differences in imaging between the 9011 and 6711 OncoSeed models. Siebert et al. (26) designed a PMMA phantom for testing the localization of nine seed models on CT scans. All the seed types were identified using the VariSeed algorithm treatment planning system (Varian Medical Systems, Inc., Palo Alto, CA) for localizing seeds, but some were placed incorrectly and this varied with seed model. The algorithm had difficulty distinguishing between closely spaced seeds, which suggests that the smaller size of the 9011 seed could be advantageous for localization using CT images.
Methods and materials We compared the visibility of the two seeds using ultrasound imaging, MRI, fluoroscopy, and CT. As our previous study had shown that there was no significant difference in the resolution (as given by the full width at half maximum [FWHM]) of seeds of similar sizes from different manufactures on any of these modalities, the 6711 OncoSeed was chosen as representative of all the seeds tested in the previous study. The exception to this was MR imaging, where the Intersource seed (Eckert & Ziegler BEBIG sa, no longer available) gave a larger signal void than the other seed types tested.
462
G. Roberts et al. / Brachytherapy 11 (2012) 460e467
Ultrasound
Magnetic resonance imaging
Phantom production Tissue equivalent prostate phantoms were made in-house using a method devised as part of previous study at our center (25). A mixture of distilled water, glycerol, agar, and cellulose (88.8%, 8.17%, 3.00%, and 0.03%, respectively, by weight) was heated to boiling point, and molded into cylindrical shapes, based on the method of Rickey et al. (27). The resulting phantom was held in a purpose made jig within a water tank and scanned using a ProFocus 12-MHz ultrasound scanner and TRUS 8848 bi-planar probe (BK Medical, Herlev, Denmark) attached to an adapted clinical cradle and needle template.
The phantoms implanted with either a single 6711 or 9011 seed under ultrasound guidance (as described above) were also used to assess visibility on MRI using a 1.5T Siemens Avanto scanner (Siemens AG, Munich, Germany). The phantoms were scanned side by side using the clinical protocol for post-implant scans: 2-mm transverse slices using a proton density weighted, turbo spin-echo pulse sequence. The slices containing the strongest signal void for each seed type were identified and profiles taken across the seeds. The relative sizes of the seeds on the images were quantified by finding the area of each signal void in pixels. Computed tomography
Needle and seed visibility on images The needle used with 6711 model seeds is a 1.27-mm diameter (18-G) needle and the needle used with 9011 seeds is a 0.91-mm needle (20-G). Needles of both gauges have beveled tips and were inserted into the phantoms using the current template. All needles were inserted in the same bevel orientation. A visual comparison between the appearance of a needle with a stylet only and needle with a seed and stylet was made. It was concluded that the two were indistinguishable, so it was decided to image the needles with stylet only. Images were taken of pairs of 18- and 20-G needles at five positions within the phantom, with each pair at a different distance from the probe. Both the needle tips and shafts (1 cm from tip) were imaged as both have clinical relevance. The positioning markings on the needle shaft were used to ensure that the needles were inserted to the same depth. The two needle types were then each loaded with a single seed and the seed inserted into separate phantoms. Images were taken of the loaded needles and of the inserted seeds with the needles removed. The needles and seeds were assessed for differences in visibility visually and by measuring the width of the echo signal on the images (perpendicular to the shadow) using the MATLAB imdistline function (MathWorks, Natick, MA).
Fluoroscopy A jig designed to hold stranded seeds was adapted to hold a row of ten 9011 seeds and one 10-seed 6711 RAPIDStrand side by side. The jig was placed in the center of a water phantom using PMMA blocks for elevation. The seeds were imaged using the Exposcop 8000 C-arm fluoroscopy unit (Ziehm Imaging GmbH, Nuremburg, Germany), which is used in theater for patient imaging. Typical clinical exposure factors of 65 kV and 2.1 mA were used during the screening and the last displayed image was digitally saved for analysis. Profiles were taken across the short axis of the seeds and averaged over the 10 seeds for each seed type.
Single seed profiles The two sets of seeds were scanned using Siemens Somatom CT scanner (Siemens AG). The same setup was used for the fluoroscopy images. A clinical prostate postimplant protocol was used: 2-mm image slice thickness, 0.9 pitch ratio, 2-mm reconstructed slice thickness with a 170 mm field of view, where 1 pixel measured 0.33 0.33 mm. The transaxial slices containing the clearest image for each of the 10 seeds were selected for analysis. Other CT orientations were not assessed because the two seed types are the same length and made of the same material so were assumed to give very similar profiles in the longitudinal direction. A MATLAB function was written to take horizontal and vertical background subtracted profiles across each seed and store the FWHM for each profile. The average size of the profiles for each seed type was compared using independent sample t-tests. As well as size, the visibility of details depends on the contrast to noise ratio (CNR). CNRs were therefore taken from the seed profiles by computing the difference between the 50% points and background and dividing this by the background. The difference in CNR was quantified using t-tests. The FWHM results were compared with those previously found for seeds from other manufacturers as part of our 2007 study (25). Resolution of closely spaced seeds One of the potential advantages of the smaller 9011 sources is that seeds that are implanted close to each other may be more easily distinguishable on CT scans, leading to more accurate post-implant dosimetry. To test how well seeds were resolved, two jigs were designed to hold pairs of seeds at varying distances apart, from 1 to 7 mm. The distances were measured from the seed centers. The jigs containing the seeds were scanned in a water tank to give realistic scatter conditions, using the same clinical protocol as above. A MATLAB function was written to take line profiles across the seed pairs and compute the degree of overlap between the profile of each seed. The overlap was measured by taking a ratio between the peak height and the minimum between the two peaks.
G. Roberts et al. / Brachytherapy 11 (2012) 460e467
463
Results
Fluoroscopy
Ultrasound
Fluoroscopy images of a series of five 6711 and 9011 seeds are shown in Fig. 2b. Line profiles taken from the average of the five seeds are shown in Fig. 3a. The width of the profile of the 9011 seed is less than the 6711 profile, but both seed types are clearly visible in the images.
There was no evidence for difference in the size of the 18and 20-G echo images at either the beveled needle tip or needle shaft. The mean differences in echo widths were 0.2 mm at the tip and 0.3 mm at 10 mm from the tip, neither of which was significant. Visually, the images of the needle tips look very similar for both needle types, but there is a slight difference in the appearance of the needle shaft because of a narrower acoustic shadow from the thinner needle, which is not seen on the images of the needle tip. The echo widths of the 18-G needle with seed and the 6711 seed only were 5.4 and 5.0 mm, respectively (35 and 32 pixels). The widths on the corresponding 9011 images were 4.8 and 3.0 mm (31 and 25 pixels). Both needle and seed types are easily visible. Example ultrasound images of both types of needles are shown in Fig. 2a.
Images of the phantoms with 6711 and 9011 seeds inserted are shown in Fig. 2c. Profiles across the seeds on transaxial slices are shown in Fig. 3b. As for CT, other orientations were not assessed. The signal void areas (measured on the slices showing the largest voids) were 22 and 12 pixels (~8.8 and 4.8 mm2) for the 6711 and 9011 seeds, respectively. The increased signal artifact surrounding the seeds in the horizontal direction is likely to be due to local distortion of the magnetic field by the metal source (28).
Fig. 2. (a) Ultrasound images of the needles and seeds in tissue-equivalent prostate phantom. Top: 6711 seed and 18-G needle; bottom: 9011 seed and 20-G needle. (b) Fluoroscopy images for the 6711 seeds (left) and 9011 seeds (right). (c) Zoomed magnetic resonance images of the 6711 seed (top) and 9011 seed (bottom), showing magnetic artifacts in the horizontal direction. The artifact is visible in the profile plots in Fig. 3 and has a greater effect on the thicker seed. (d) Zoomed inverted computed tomography images of the 6711 seed (top) and 9011 seed (bottom).
Fig. 3. Line profiles taken across seeds on a fluoroscopy frame (a), magnetic resonance image central slice (b), and computed tomography slice (c). The difference between the 9011 seed profile and 6711 seed profile is greatest on the magnetic resonance image.
Magnetic resonance imaging
464
G. Roberts et al. / Brachytherapy 11 (2012) 460e467
Computed tomography Single seed profiles Example CT slice images of the seeds taken with the clinical protocol are shown in Fig. 2d. Line profiles averaged from vertical and horizontal profiles of ten 6711 and 9011 seeds are shown in Fig. 3c. The 6711 seeds generate more reconstruction (streak) artifacts than the thinner seeds. The FWHM of the profiles was significantly narrower for the 9011 seeds than the 6711 seeds (1.42 c.f. 1.92 mm, p! 0.001). The CNR of the 9011 seeds was slightly lower (2.39 c.f. 2.45). This was statistically significant for some but not all of the profiles taken, indicating a negligible difference. Figure 4 shows the FWM 50, FWM 75, and FWM 80 measurements, relative to the OncoSeed 6711 model, for the 9011 model and for the other seeds tested in our previous study (25). These FWM (full width maximum) measurements are the profile widths at 50%, 75%, and 80% of the maximum height, respectively.
Resolution CT images of the seeds held fixed distances apart are shown for the 6711 and 9011 models in Fig. 5a. The reduction in artifacts with the narrower seeds is more apparent on these images of multiple seeds. These are caused by the back projection of the signal from the high contrast seeds during image reconstruction. Due to the seed arrangement, there is a greater degree of image artifact in the horizontal direction than in the vertical direction. For this reason, vertical profiles only were used for analysis. A figure of merit showing the percentage overlap of the profiles for each seed is shown in Fig. 5b. The overlap was measured by taking a ratio between the peak height and the minimum between the two peaks from the adjacent seeds. Plots showing profiles of pairs of seeds at a separation of 2 mm are shown in Fig. 6. We found that the 9011 seeds are resolved better than the 6711 seeds, especially at
separations of 2e3 mm. There is little difference at 1-mm separation, where neither seed model is well resolved.
Discussion Ultrasound Although there was no evidence for differences in the size of the two needle types on ultrasound images, it should be noted that because of the flexibility of the 9011 needle and the relatively large holes in the current template it was difficult to ensure that pairs of needles were always inserted at exactly the same distance from the ultrasound probe. Our previous analysis (Al-Qaisieh et al. (25)) showed that visibility of the seeds on ultrasound images is highly dependent on the orientation of the seed relative to the TRUS probe and that this factor outweighs any interseed differences. The distance between the seed and ultrasound probe affects the size of the echo seen on the images. For the images of the seeds being inserted, the needles were successfully placed at the same distance from the ultrasound probe. The measurements show the signal from the 9011 needle and seed is slightly smaller. Visually, the needles give similar images, although the shadow of the smaller needle appears weaker. At SJIO, only one needle is inserted at a time in clinical practice so this is not significant. However at centers where more than one needle is imaged simultaneously, the reduced acoustic shadow could be advantageous. The flexibility of the thinner needles could prove a disadvantage when implanting seeds because the planning system initially assumes they are inserted straight and parallel. Interactive treatment planning overcomes this disadvantage by allowing the inserted needle projection to be taken into account. This will need to be evaluated using a template designed specifically for 20-G needles. Although it was generally possible to tell the needles apart, radiologists who viewed the images agreed that both
Fig. 4. The full width at half maximum and full width at 75% and 80% for profiles for Oncura OncoSeed 6711, 9011, and Echoseed 6733; Bebig IsoSeed and Intersource; and Nucletron SelectSeed on transverse computed tomography images. The results for the seeds other than Oncura OncoSeed were obtained in a previous study (Al-Qaisieh et al. (25)) undertaken at our center. Horizontal profiles are shown in (a) and vertical in (b).
G. Roberts et al. / Brachytherapy 11 (2012) 460e467
465
Fluoroscopy The images in Fig. 2 show that both seed types were easily visible on the fluoroscopy images taken using clinical settings and scattering medium. The plots in Fig. 3 show that the profile of the model 9011 seed is narrower than the 6711 profile, but of a similar height (contrast). The seeds give such high levels of contrast on X-ray images, as a result of the titanium encapsulation and gold marker, that the differences in seed size have a negligible effect on how well they can be detected. The narrower profile of the 9011 seed may be advantageous for resolving seeds on fluoroscopy images used for checking seed placement immediately after implant.
Magnetic resonance imaging
Fig. 5. Computed tomography images (inverted) of the two phantoms holding 6711 seeds (a) and 9011 seeds (b) at different distances apart. The artifacts are clearly less obvious in the bottom image and the images of the seeds are also less distorted. The accompanying plot shows the degree of overlap of adjacent vertical profiles for the two seed types for each seed separation distance. The profiles from the thinner 9011 seed are better resolved (less overlap) at all separations.
types of needle were equally well visible on the images and that either would be clinically acceptable. There was substantial intraneedle variability in appearance because of positioning within the phantom and orientation with respect to the probe, and this outweighed the variability between the needle types.
Our previous analysis (Al-Qaisieh et al. (25)) showed that signal void measurements from the MRI images for various seed types with the exception of Intersource were consistent in showing a signal void of 4 mm for all scanning protocols. Intersource showed a consistent signal void of 9 mm for all scanning protocols. The present study shows the maximum signal void for the 9011 seed model was approximately half the size of the 6711 void. The significance of this in terms of seed identification for post-plan dosimetry will depend on the size and appearance of other structures within the patient that give rise to a loss in MR signal, such as blood vessels. However, the artifact seen on the images could, if reproducible, aid in distinguishing seeds from other structures. At our center, MR imaging is always used in conjunction with CT data and so the visibility of the 9011 seeds on MRI is not critical. However, in centers where MR images alone are used for post-plan dosimetry, the visibility of the seeds would need to be investigated in a clinical context. This could be done by measuring the size of signal voids due to 6711 seeds and other structures on a sample of previous patient images, and assuming that the 9011 seeds would appear at half the size of the 6711 seeds in the axial plane. It would also be useful to investigate whether the smaller signal voids of the 9011 model seeds might improve the delineation of the prostate.
Fig. 6. Plots of pairs of seed profiles with 2-mm separation between seeds for 6711 (RapidStrand) (a) and 9011 (ThinSeed) (b). The plots show a clear improvement in resolution with the 9011 seeds.
466
G. Roberts et al. / Brachytherapy 11 (2012) 460e467
Computed tomography
References
Figure 4 shows that the 9011 seeds give a lower signal (as measured by the FWHM of line profiles across the seeds) than all the seeds previously tested. However, the narrower profile and lower signal to noise ratio of the 9011 seed would not be expected to have a clinically significant effect on how well the seeds could be detected, assuming the clinical CT images obtained are of good image quality. Furthermore, because both seed types are clearly visible on CT images, the reduction in artifacts generated by the thinner seeds would be advantageous for seed identification, especially where multiple seeds are present in the same transverse slice. As other seeds have been shown to give similar CT signals as the 6711 seeds (25), it can be assumed that the 9011 seeds would also give a reduction in CT artifacts when compared with other models. The thinner 9011 seeds were resolved better than the 6711 seeds at close separations, particularly below 3 mm. The seeds at our center are not usually planned to be implanted at smaller separations than 7 mm, but often appear closer on the postplan scan because of uncertainty in the implant process and postimplant edema. The use of 9011 seeds may therefore also improve the seed detection accuracy performed on CTand aid the automatic seed detection algorithm. The VariSeed automatic seed recognition algorithm Seedfinder (Varian Medical Systems Inc., Palo Alto, CA) is currently configured for locating 6711 seeds. It would be advisable to validate Seedfinder with 9011 seeds before clinical use, using a reproducible rigid seed arrangement similar to the phantom produced by Siebert et al. (26) or De Brabandere et al. (29). Future algorithms could potentially incorporate seed profile recognition into the detection system, where the narrower profiles and reduction in artifacts of the 9011 seeds would be an advantage. A reduction in seed artifacts could also prove to be advantageous when using dose calculation techniques such as Monte Carlo.
[1] Salembier C, Lavagnini P, Nickers P, et al. Tumour and target volumes in permanent prostate brachytherapy: A supplement to the ESTRO/EAU/EORTC recommendations on prostate brachytherapy. Radiother Oncol 2007;83:3e10. [2] Ash D, Al-Qaisieh B, Bottomley D, et al. The correlation between D90 and outcome for I-125 seed implant monotherapy for localised prostate cancer. Radiother Oncol 2006;79:185e189. [3] Potters L, Cao Y, Calugaru E. 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. [4] Stock RG, Stone NN, Tarbert A, et al. A dose response study for I-125 prostate implants. Int J Radiat Oncol Biol Phys 1998;41: 101e108. [5] Rivard MJ. Monte Carlo radiation dose simulations and dosimetric comparison of the model 6711 and 9011 I-125 brachytherapy sources. Med Phys 2009;36:486e491. [6] Kennedy RM, Davis SD, Micka JA, et al. Experimental and Monte Carlo determination of the TG-43 dosimetruc parameters for the model 9011 THINSeedÔ brachytherapy source. Med Phys 2010;37: 1681e1688. [7] Keyes M, Miller S, Moravan V, et al. Predictive factors for acute and late urinary toxicity after permanent prostate brachytherapy: Long term outcome in 712 consecutive patients. Int J Radiat Oncol Biol Phys 2009;73:1023e1032. [8] Nag S, Bice W, Dewyngaert K, et al. The American Brachytherapy Society recommendations for permanent prostate brachytherapy postimplant dosimetric analysis. Int J Radiat Oncol Biol Phys 2000;46: 221e230. [9] Yue N, Mori J, Nath R, et al. External beam radiotherapy boosts to reduce the impact caused by edema in prostate permanent seed implants. Phys Med Biol 2006;51:2267e2277. [10] Crook J, Mclean M, Yeung I. MRI-CT fusion to assess postbrachytherapy prostate volume and the effects of prolonged edema on dosimetry following transperineal interstitial permanent prostate brachytherapy. Brachytherapy 2004;3:55e60. [11] Taussky D, Austen L, Toi A, et al. Sequential evaluation of prostate edema after permanent seed prostate brachytherapy using CT-MRI fusion. Int J Radiat Oncol Biol Phys 2005;62:974e980. [12] Pinkawa M, Asadpour B, Gagel B, et al. Evaluation of source displacement and doseevolume changes after permanent prostate brachytherapy with stranded seeds. Radiother Oncol 2007;84: 190e196. [13] Ali I, Algan O, Thompson S, et al. A comparative study of seed localization and dose calculation on pre and post-implantation ultrasound and CT images for low-dose-rate prostate brachytherapy. Phys Med Biol 2009;54:5595e5611. [14] Papagikos MA, Deguzman AF, Rossi PJ, et al. Dosimetric quantifiers for low-dose-rate prostate brachytherapy: Is V100 superior to D90? Brachytherapy 2005;4:252e258. [15] Miles EF, Nelson JW, Alkaissi AK, et al. Equivalent uniform dose, D90, and V100 correlation with biochemical control after lowdose-rate prostate brachytherapy for clinically low-risk prostate cancer. Brachytherapy 2008;7:206e211. [16] Buskirk SJ, Pinkstaff DM, Petrou SP, et al. Acute urinary retention after transperineal template guided prostate biopsy. Int J Radiat Oncol Biol Phys 2004;59:1360e1366. [17] Stone NM, Stock RG. Complications following permanent prostate brachytherapy. Eur Urol 2002;41:427e433. [18] Eapen L, Kayser C, Deshaies Y, et al. Correlating the degree of needle trauma during prostate brachytherapy and the development of urinary toxicity. Int J Radiat Oncol Biol Phys 2004;59:1392e1394. [19] Lee N, Wuu CS, Brody R, et al. Factors predicting for postimplantation urinary retention after permanent prostate brachytherapy. Int J Radiat Oncol Biol Phys 2000;48:1457e1460.
Conclusion The visibility of the 9011 seeds and/or needles on ultrasound, fluoroscopy, and CT images is clinically acceptable. Use of the narrower seeds is likely to improve CT post-plan dosimetry by reducing artifacts and aiding the identification of closely spaced seeds. Further study is required into the suitability of MR images alone for post-plan dosimetry using the 9011 model seeds.
Acknowledgments The authors acknowledge radiologists Dr. Jonathan Smith and Dr. Brendan Carey for their assistance in the assessment of the clinical acceptability of the fluoroscopy images acquired for this work.
G. Roberts et al. / Brachytherapy 11 (2012) 460e467 [20] Bucci J, Morris WJ, Keyes M, et al. Predictive factors of urinary retention following prostate brachytherapy. Int J Radiat Oncol Biol Phys 2002;53:91e98. [21] Neill M, Studer G, Le L, et al. The nature and extent of urinary morbidity in relation to prostate brachytherapy urethral dosimetry. Brachytherapy 2006;6:173e179. [22] Wong J, Sylvester JE, Grimm PD, et al. First report on the use of a thinner I-125 radioactive seed within 20 gauge needles for permanent seed prostate brachytherapy: Evaluation of post-implant dosimetry and acute toxicity. Int J Radiat Oncol Biol Phys 2009; 75:S325. [23] Heintz BH, Wallace RE, Hevezi JM. Comparison of I-125 sources used for permanent interstitial implant. Med Phys 2001;28:671e682. [24] Meigooni AS, Luerman CM, Sowards KT. Evaluation of the dose distribution for prostate implants using various I-125 and Pd-103 sources. Med Phys 2009;36:2267e2277.
467
[25] Al-Qaisieh B, Smith DW, Brearley E, et al. Comprehensive I-125 multi-seed comparison for prostate brachytherapy: Dosimetry and visibility analysis. Radiother Oncol 2007;84:140e147. [26] Siebert FA, De Brabandere M, Kirisits C, et al. Phantom investigations on CT seed imaging for interstitial brachytherapy. Radiother Oncol 2007;85:316e323. [27] Rickey DW, Picot PA, Christopher D, et al. A wall-less vessel phantom for Doppler ultrasound studies. Ultrasound Med Biol 1995;21: 1163e1176. [28] Haacke EM, Brown RW, Thompson MR, et al. Magnetic field inhomogeneity effects and T2 dephasing. In: Haacke EM, editor. Magnetic resonance imaging: physical principles and sequence design. New York, NY: Wiley-Liss; 1999. p. 570e580. [29] De Brabandere M, Kirisitis C, Peeters R, et al. Accuracy of seed reconstruction in prostate postplanning studied with a CT- and MRI-compatible phantom. Radiother Oncol 2006;79:190e197.
Evaluation of the visibility of a new thinner 125I radioactive source for permanent prostate brachytherapy Gemma Roberts, Bashar Al-Qaisieh*, Peter Bownes Medical Physics and Engineering, St James’s Institute of Oncology, St James’s Hospital, Leeds, UK
ABSTRACT
PURPOSE: The 125I source currently used for prostate brachytherapy at St. James’s Institute of Oncology is a standard size seed (z4.5 mm in length and 0.8 mm in diameter). A new, thinner seed is under evaluation. This is designed to be implanted using narrower needles, potentially reducing edema and improving the dose distribution. This study investigated the visibility of the thinner source on multimodality images and compared it with that of standard size seeds. METHODS AND MATERIALS: Images of dummy seeds of both thinner and standard size models were taken using ultrasound, fluoroscopy, computed tomography (CT), and magnetic resonance (MR) imaging. The ultrasound, fluoroscopy, and CT images were acquired with the seeds inserted into phantoms positioned in a water tank. The MR images were acquired using phantoms containing single seeds. The images were analyzed visually and quantitatively. The resolution of closely spaced seeds on CT images was investigated. RESULTS: The visibility of both seeds was similar on ultrasound, fluoroscopy, and MR images. On CT images, the thinner seeds give reduced artifacts and better resolution. CONCLUSIONS: The use of the thinner seed would have minimal effect on ultrasound and fluoroscopy imaging during treatment. However on CT images, the use of the thinner seeds may improve seed identification for post-treatment dosimetry. Further study is required into the suitability of MR images alone for post-treatment dosimetry. Ó 2012 American Brachytherapy Society. Published by Elsevier Inc. All rights reserved.
Keywords:
Prostate brachytherapy;
125
I; OncoSeed Model 9011; THIN Seed; Visibility
Background and purpose Early stage prostate cancer may be treated with permanent brachytherapy using 125I sources (1). The seed currently used at St. James’ Institute of Oncology (SJIO) is the OncoSeed model 6711 (2) (Oncura Inc., Arlington Heights, IL). The seeds are supplied sutured into strands of 10 (RAPIDStrand; Oncura Inc.), which can be cut to size. OncoSeed model 6711 is used in many clinical reviews of treatment outcome currently available (3, 4). The OncoSeed model 6711 source Received 23 September 2011; received in revised form 12 January 2012; accepted 12 January 2012. Financial disclosure: The authors above have no conflict of interest in connection with the article and the material described is not under publication or consideration for publication elsewhere. The non-radioactive dummy seeds for this imaging study were provided by Oncura Inc., Arlington Heights, IL). The study was designed and carried out independently of Oncura Inc. * Corresponding author. St James’s Institute of Oncology, St James’s Hospital, Beckett Street, Leeds LS9 7TF, UK. Tel.: þ44-113-206-7409. E-mail address: [email protected] (B. Al-Qaisieh).
consists of an active core of 125I adsorbed onto the surface of a silver rod (length 2.8 mm, diameter 0.508 mm) contained within a titanium capsule (length 4.56 mm, diameter 0.774 mm) (5). The OncoSeed model 6711 seeds are implanted using 18-G needles, which have an outer diameter of 1.27 mm. A new, thinner seed model (OncoSeed model 9011dTHIN Seed; Oncura Inc., Amersham, UK) has recently been developed and our center has obtained some non-radioactive dummy seeds for evaluation. The OncoSeed 9011 model seed is of a similar design to the 6711 and is manufactured in a similar way (6). The OncoSeed model 9011 seed will also be available in sutured strands (THIN Strand; Oncura Inc.). The main difference is the smaller diameter (source 0.30 mm, capsule 0.508 mm) (Fig. 1), which allows narrower 20-G needles (0.91 mm outer diameter) to be used during the implant process. This will potentially reduce the level of edema and trauma to the prostate and urethra (7). The prostate volume may change during and after the implant process because of edema (8). Several studies have
1538-4721/$ - see front matter Ó 2012 American Brachytherapy Society. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.brachy.2012.01.009
G. Roberts et al. / Brachytherapy 11 (2012) 460e467
461
Fig. 1. Cross-sectional diagram showing the dimensions of the OncoSeed 6711 and OncoSeed 9011 (THIN Seed) (Image courtesy of Oncura Inc.).
shown that edema can affect the percentage of the prostate receiving the prescribed dose (V100) by moving the seeds and increasing the volume (9e13). V100 and D90 (the dose received by 90% of the prostate) have been linked to patient outcome (3, 4, 14, 15). Assuming equivalent dosimetry, the new seed model could therefore potentially improve the dose coverage to the prostate by reducing the level of edema. The study of Buskirk et al. (16) suggests that needle trauma is the main cause of short-term urinary symptoms. Several other studies (17e22) have found a statistically significant correlation ( p 5 0.025) (18) between needle trauma during implant (via the number of needles) and edema and urinary retention, which suggests that the new narrower needles could prove advantageous. At the Seattle Prostate Institute, Seattle, WA, model 9011 OncoSeeds have been in clinical use since August 2008 (22). Initial results based on the first 100 patients treated have shown improved results for post-treatment dosimetry when compared with previous patients treated with model 6711 seeds, similar urinary retention rates, and lower levels of pain and bruising. Several imaging modalities are used in the treatment and followup of 125I prostate brachytherapy patients. Transrectal ultrasound (TRUS) is used for guiding needle insertion and in some centers for imaging seeds after implant (23). Fluoroscopy screening can be used at various stages during implant to check the positioning of the seeds. Computed tomography (CT) and magnetic resonance imaging (MRI) scans are used to identify seed positions for post-implant dosimetry. Before bringing a new source into clinical use, it is important to ensure that images obtained with the source are fit for purpose. This study aims to assess whether the introduction of model 9011 seeds and the associated 20-G narrow bore needle will have a clinically significant impact on the appearance and localization of the seeds using any of the above imaging techniques. The importance of source visibility is mentioned by Heintz et al. (23) and Meigooni et al. (24) 2009, but
only a small number of studies characterizing prostate brachytherapy sources take account of this aspect. Previous work at this center by Al-Qaisieh et al. (25) compared the visibility of five seeds implanted in a tissue equivalent phantom and imaged using ultrasound, CT, MRI, and fluoroscopy. Clinically significant differences in visibility were not found in this study. The present study is a continuation of the previous work by Al-Qaisieh et al. and was initiated as a result of the new THIN Seed source coming to market. The two studies jointly represent a comprehensive review of the visibility of 125 I sources from some of the more common manufacturers in Europe (Oncura Inc., Nucletron B.V, Veenendaal, Netherlands and Eckert & Ziegler BEBIG sa, Seneffe, Belgium). This is the first study looking at the differences in imaging between the 9011 and 6711 OncoSeed models. Siebert et al. (26) designed a PMMA phantom for testing the localization of nine seed models on CT scans. All the seed types were identified using the VariSeed algorithm treatment planning system (Varian Medical Systems, Inc., Palo Alto, CA) for localizing seeds, but some were placed incorrectly and this varied with seed model. The algorithm had difficulty distinguishing between closely spaced seeds, which suggests that the smaller size of the 9011 seed could be advantageous for localization using CT images.
Methods and materials We compared the visibility of the two seeds using ultrasound imaging, MRI, fluoroscopy, and CT. As our previous study had shown that there was no significant difference in the resolution (as given by the full width at half maximum [FWHM]) of seeds of similar sizes from different manufactures on any of these modalities, the 6711 OncoSeed was chosen as representative of all the seeds tested in the previous study. The exception to this was MR imaging, where the Intersource seed (Eckert & Ziegler BEBIG sa, no longer available) gave a larger signal void than the other seed types tested.
462
G. Roberts et al. / Brachytherapy 11 (2012) 460e467
Ultrasound
Magnetic resonance imaging
Phantom production Tissue equivalent prostate phantoms were made in-house using a method devised as part of previous study at our center (25). A mixture of distilled water, glycerol, agar, and cellulose (88.8%, 8.17%, 3.00%, and 0.03%, respectively, by weight) was heated to boiling point, and molded into cylindrical shapes, based on the method of Rickey et al. (27). The resulting phantom was held in a purpose made jig within a water tank and scanned using a ProFocus 12-MHz ultrasound scanner and TRUS 8848 bi-planar probe (BK Medical, Herlev, Denmark) attached to an adapted clinical cradle and needle template.
The phantoms implanted with either a single 6711 or 9011 seed under ultrasound guidance (as described above) were also used to assess visibility on MRI using a 1.5T Siemens Avanto scanner (Siemens AG, Munich, Germany). The phantoms were scanned side by side using the clinical protocol for post-implant scans: 2-mm transverse slices using a proton density weighted, turbo spin-echo pulse sequence. The slices containing the strongest signal void for each seed type were identified and profiles taken across the seeds. The relative sizes of the seeds on the images were quantified by finding the area of each signal void in pixels. Computed tomography
Needle and seed visibility on images The needle used with 6711 model seeds is a 1.27-mm diameter (18-G) needle and the needle used with 9011 seeds is a 0.91-mm needle (20-G). Needles of both gauges have beveled tips and were inserted into the phantoms using the current template. All needles were inserted in the same bevel orientation. A visual comparison between the appearance of a needle with a stylet only and needle with a seed and stylet was made. It was concluded that the two were indistinguishable, so it was decided to image the needles with stylet only. Images were taken of pairs of 18- and 20-G needles at five positions within the phantom, with each pair at a different distance from the probe. Both the needle tips and shafts (1 cm from tip) were imaged as both have clinical relevance. The positioning markings on the needle shaft were used to ensure that the needles were inserted to the same depth. The two needle types were then each loaded with a single seed and the seed inserted into separate phantoms. Images were taken of the loaded needles and of the inserted seeds with the needles removed. The needles and seeds were assessed for differences in visibility visually and by measuring the width of the echo signal on the images (perpendicular to the shadow) using the MATLAB imdistline function (MathWorks, Natick, MA).
Fluoroscopy A jig designed to hold stranded seeds was adapted to hold a row of ten 9011 seeds and one 10-seed 6711 RAPIDStrand side by side. The jig was placed in the center of a water phantom using PMMA blocks for elevation. The seeds were imaged using the Exposcop 8000 C-arm fluoroscopy unit (Ziehm Imaging GmbH, Nuremburg, Germany), which is used in theater for patient imaging. Typical clinical exposure factors of 65 kV and 2.1 mA were used during the screening and the last displayed image was digitally saved for analysis. Profiles were taken across the short axis of the seeds and averaged over the 10 seeds for each seed type.
Single seed profiles The two sets of seeds were scanned using Siemens Somatom CT scanner (Siemens AG). The same setup was used for the fluoroscopy images. A clinical prostate postimplant protocol was used: 2-mm image slice thickness, 0.9 pitch ratio, 2-mm reconstructed slice thickness with a 170 mm field of view, where 1 pixel measured 0.33 0.33 mm. The transaxial slices containing the clearest image for each of the 10 seeds were selected for analysis. Other CT orientations were not assessed because the two seed types are the same length and made of the same material so were assumed to give very similar profiles in the longitudinal direction. A MATLAB function was written to take horizontal and vertical background subtracted profiles across each seed and store the FWHM for each profile. The average size of the profiles for each seed type was compared using independent sample t-tests. As well as size, the visibility of details depends on the contrast to noise ratio (CNR). CNRs were therefore taken from the seed profiles by computing the difference between the 50% points and background and dividing this by the background. The difference in CNR was quantified using t-tests. The FWHM results were compared with those previously found for seeds from other manufacturers as part of our 2007 study (25). Resolution of closely spaced seeds One of the potential advantages of the smaller 9011 sources is that seeds that are implanted close to each other may be more easily distinguishable on CT scans, leading to more accurate post-implant dosimetry. To test how well seeds were resolved, two jigs were designed to hold pairs of seeds at varying distances apart, from 1 to 7 mm. The distances were measured from the seed centers. The jigs containing the seeds were scanned in a water tank to give realistic scatter conditions, using the same clinical protocol as above. A MATLAB function was written to take line profiles across the seed pairs and compute the degree of overlap between the profile of each seed. The overlap was measured by taking a ratio between the peak height and the minimum between the two peaks.
G. Roberts et al. / Brachytherapy 11 (2012) 460e467
463
Results
Fluoroscopy
Ultrasound
Fluoroscopy images of a series of five 6711 and 9011 seeds are shown in Fig. 2b. Line profiles taken from the average of the five seeds are shown in Fig. 3a. The width of the profile of the 9011 seed is less than the 6711 profile, but both seed types are clearly visible in the images.
There was no evidence for difference in the size of the 18and 20-G echo images at either the beveled needle tip or needle shaft. The mean differences in echo widths were 0.2 mm at the tip and 0.3 mm at 10 mm from the tip, neither of which was significant. Visually, the images of the needle tips look very similar for both needle types, but there is a slight difference in the appearance of the needle shaft because of a narrower acoustic shadow from the thinner needle, which is not seen on the images of the needle tip. The echo widths of the 18-G needle with seed and the 6711 seed only were 5.4 and 5.0 mm, respectively (35 and 32 pixels). The widths on the corresponding 9011 images were 4.8 and 3.0 mm (31 and 25 pixels). Both needle and seed types are easily visible. Example ultrasound images of both types of needles are shown in Fig. 2a.
Images of the phantoms with 6711 and 9011 seeds inserted are shown in Fig. 2c. Profiles across the seeds on transaxial slices are shown in Fig. 3b. As for CT, other orientations were not assessed. The signal void areas (measured on the slices showing the largest voids) were 22 and 12 pixels (~8.8 and 4.8 mm2) for the 6711 and 9011 seeds, respectively. The increased signal artifact surrounding the seeds in the horizontal direction is likely to be due to local distortion of the magnetic field by the metal source (28).
Fig. 2. (a) Ultrasound images of the needles and seeds in tissue-equivalent prostate phantom. Top: 6711 seed and 18-G needle; bottom: 9011 seed and 20-G needle. (b) Fluoroscopy images for the 6711 seeds (left) and 9011 seeds (right). (c) Zoomed magnetic resonance images of the 6711 seed (top) and 9011 seed (bottom), showing magnetic artifacts in the horizontal direction. The artifact is visible in the profile plots in Fig. 3 and has a greater effect on the thicker seed. (d) Zoomed inverted computed tomography images of the 6711 seed (top) and 9011 seed (bottom).
Fig. 3. Line profiles taken across seeds on a fluoroscopy frame (a), magnetic resonance image central slice (b), and computed tomography slice (c). The difference between the 9011 seed profile and 6711 seed profile is greatest on the magnetic resonance image.
Magnetic resonance imaging
464
G. Roberts et al. / Brachytherapy 11 (2012) 460e467
Computed tomography Single seed profiles Example CT slice images of the seeds taken with the clinical protocol are shown in Fig. 2d. Line profiles averaged from vertical and horizontal profiles of ten 6711 and 9011 seeds are shown in Fig. 3c. The 6711 seeds generate more reconstruction (streak) artifacts than the thinner seeds. The FWHM of the profiles was significantly narrower for the 9011 seeds than the 6711 seeds (1.42 c.f. 1.92 mm, p! 0.001). The CNR of the 9011 seeds was slightly lower (2.39 c.f. 2.45). This was statistically significant for some but not all of the profiles taken, indicating a negligible difference. Figure 4 shows the FWM 50, FWM 75, and FWM 80 measurements, relative to the OncoSeed 6711 model, for the 9011 model and for the other seeds tested in our previous study (25). These FWM (full width maximum) measurements are the profile widths at 50%, 75%, and 80% of the maximum height, respectively.
Resolution CT images of the seeds held fixed distances apart are shown for the 6711 and 9011 models in Fig. 5a. The reduction in artifacts with the narrower seeds is more apparent on these images of multiple seeds. These are caused by the back projection of the signal from the high contrast seeds during image reconstruction. Due to the seed arrangement, there is a greater degree of image artifact in the horizontal direction than in the vertical direction. For this reason, vertical profiles only were used for analysis. A figure of merit showing the percentage overlap of the profiles for each seed is shown in Fig. 5b. The overlap was measured by taking a ratio between the peak height and the minimum between the two peaks from the adjacent seeds. Plots showing profiles of pairs of seeds at a separation of 2 mm are shown in Fig. 6. We found that the 9011 seeds are resolved better than the 6711 seeds, especially at
separations of 2e3 mm. There is little difference at 1-mm separation, where neither seed model is well resolved.
Discussion Ultrasound Although there was no evidence for differences in the size of the two needle types on ultrasound images, it should be noted that because of the flexibility of the 9011 needle and the relatively large holes in the current template it was difficult to ensure that pairs of needles were always inserted at exactly the same distance from the ultrasound probe. Our previous analysis (Al-Qaisieh et al. (25)) showed that visibility of the seeds on ultrasound images is highly dependent on the orientation of the seed relative to the TRUS probe and that this factor outweighs any interseed differences. The distance between the seed and ultrasound probe affects the size of the echo seen on the images. For the images of the seeds being inserted, the needles were successfully placed at the same distance from the ultrasound probe. The measurements show the signal from the 9011 needle and seed is slightly smaller. Visually, the needles give similar images, although the shadow of the smaller needle appears weaker. At SJIO, only one needle is inserted at a time in clinical practice so this is not significant. However at centers where more than one needle is imaged simultaneously, the reduced acoustic shadow could be advantageous. The flexibility of the thinner needles could prove a disadvantage when implanting seeds because the planning system initially assumes they are inserted straight and parallel. Interactive treatment planning overcomes this disadvantage by allowing the inserted needle projection to be taken into account. This will need to be evaluated using a template designed specifically for 20-G needles. Although it was generally possible to tell the needles apart, radiologists who viewed the images agreed that both
Fig. 4. The full width at half maximum and full width at 75% and 80% for profiles for Oncura OncoSeed 6711, 9011, and Echoseed 6733; Bebig IsoSeed and Intersource; and Nucletron SelectSeed on transverse computed tomography images. The results for the seeds other than Oncura OncoSeed were obtained in a previous study (Al-Qaisieh et al. (25)) undertaken at our center. Horizontal profiles are shown in (a) and vertical in (b).
G. Roberts et al. / Brachytherapy 11 (2012) 460e467
465
Fluoroscopy The images in Fig. 2 show that both seed types were easily visible on the fluoroscopy images taken using clinical settings and scattering medium. The plots in Fig. 3 show that the profile of the model 9011 seed is narrower than the 6711 profile, but of a similar height (contrast). The seeds give such high levels of contrast on X-ray images, as a result of the titanium encapsulation and gold marker, that the differences in seed size have a negligible effect on how well they can be detected. The narrower profile of the 9011 seed may be advantageous for resolving seeds on fluoroscopy images used for checking seed placement immediately after implant.
Magnetic resonance imaging
Fig. 5. Computed tomography images (inverted) of the two phantoms holding 6711 seeds (a) and 9011 seeds (b) at different distances apart. The artifacts are clearly less obvious in the bottom image and the images of the seeds are also less distorted. The accompanying plot shows the degree of overlap of adjacent vertical profiles for the two seed types for each seed separation distance. The profiles from the thinner 9011 seed are better resolved (less overlap) at all separations.
types of needle were equally well visible on the images and that either would be clinically acceptable. There was substantial intraneedle variability in appearance because of positioning within the phantom and orientation with respect to the probe, and this outweighed the variability between the needle types.
Our previous analysis (Al-Qaisieh et al. (25)) showed that signal void measurements from the MRI images for various seed types with the exception of Intersource were consistent in showing a signal void of 4 mm for all scanning protocols. Intersource showed a consistent signal void of 9 mm for all scanning protocols. The present study shows the maximum signal void for the 9011 seed model was approximately half the size of the 6711 void. The significance of this in terms of seed identification for post-plan dosimetry will depend on the size and appearance of other structures within the patient that give rise to a loss in MR signal, such as blood vessels. However, the artifact seen on the images could, if reproducible, aid in distinguishing seeds from other structures. At our center, MR imaging is always used in conjunction with CT data and so the visibility of the 9011 seeds on MRI is not critical. However, in centers where MR images alone are used for post-plan dosimetry, the visibility of the seeds would need to be investigated in a clinical context. This could be done by measuring the size of signal voids due to 6711 seeds and other structures on a sample of previous patient images, and assuming that the 9011 seeds would appear at half the size of the 6711 seeds in the axial plane. It would also be useful to investigate whether the smaller signal voids of the 9011 model seeds might improve the delineation of the prostate.
Fig. 6. Plots of pairs of seed profiles with 2-mm separation between seeds for 6711 (RapidStrand) (a) and 9011 (ThinSeed) (b). The plots show a clear improvement in resolution with the 9011 seeds.
466
G. Roberts et al. / Brachytherapy 11 (2012) 460e467
Computed tomography
References
Figure 4 shows that the 9011 seeds give a lower signal (as measured by the FWHM of line profiles across the seeds) than all the seeds previously tested. However, the narrower profile and lower signal to noise ratio of the 9011 seed would not be expected to have a clinically significant effect on how well the seeds could be detected, assuming the clinical CT images obtained are of good image quality. Furthermore, because both seed types are clearly visible on CT images, the reduction in artifacts generated by the thinner seeds would be advantageous for seed identification, especially where multiple seeds are present in the same transverse slice. As other seeds have been shown to give similar CT signals as the 6711 seeds (25), it can be assumed that the 9011 seeds would also give a reduction in CT artifacts when compared with other models. The thinner 9011 seeds were resolved better than the 6711 seeds at close separations, particularly below 3 mm. The seeds at our center are not usually planned to be implanted at smaller separations than 7 mm, but often appear closer on the postplan scan because of uncertainty in the implant process and postimplant edema. The use of 9011 seeds may therefore also improve the seed detection accuracy performed on CTand aid the automatic seed detection algorithm. The VariSeed automatic seed recognition algorithm Seedfinder (Varian Medical Systems Inc., Palo Alto, CA) is currently configured for locating 6711 seeds. It would be advisable to validate Seedfinder with 9011 seeds before clinical use, using a reproducible rigid seed arrangement similar to the phantom produced by Siebert et al. (26) or De Brabandere et al. (29). Future algorithms could potentially incorporate seed profile recognition into the detection system, where the narrower profiles and reduction in artifacts of the 9011 seeds would be an advantage. A reduction in seed artifacts could also prove to be advantageous when using dose calculation techniques such as Monte Carlo.
[1] Salembier C, Lavagnini P, Nickers P, et al. Tumour and target volumes in permanent prostate brachytherapy: A supplement to the ESTRO/EAU/EORTC recommendations on prostate brachytherapy. Radiother Oncol 2007;83:3e10. [2] Ash D, Al-Qaisieh B, Bottomley D, et al. The correlation between D90 and outcome for I-125 seed implant monotherapy for localised prostate cancer. Radiother Oncol 2006;79:185e189. [3] Potters L, Cao Y, Calugaru E. 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. [4] Stock RG, Stone NN, Tarbert A, et al. A dose response study for I-125 prostate implants. Int J Radiat Oncol Biol Phys 1998;41: 101e108. [5] Rivard MJ. Monte Carlo radiation dose simulations and dosimetric comparison of the model 6711 and 9011 I-125 brachytherapy sources. Med Phys 2009;36:486e491. [6] Kennedy RM, Davis SD, Micka JA, et al. Experimental and Monte Carlo determination of the TG-43 dosimetruc parameters for the model 9011 THINSeedÔ brachytherapy source. Med Phys 2010;37: 1681e1688. [7] Keyes M, Miller S, Moravan V, et al. Predictive factors for acute and late urinary toxicity after permanent prostate brachytherapy: Long term outcome in 712 consecutive patients. Int J Radiat Oncol Biol Phys 2009;73:1023e1032. [8] Nag S, Bice W, Dewyngaert K, et al. The American Brachytherapy Society recommendations for permanent prostate brachytherapy postimplant dosimetric analysis. Int J Radiat Oncol Biol Phys 2000;46: 221e230. [9] Yue N, Mori J, Nath R, et al. External beam radiotherapy boosts to reduce the impact caused by edema in prostate permanent seed implants. Phys Med Biol 2006;51:2267e2277. [10] Crook J, Mclean M, Yeung I. MRI-CT fusion to assess postbrachytherapy prostate volume and the effects of prolonged edema on dosimetry following transperineal interstitial permanent prostate brachytherapy. Brachytherapy 2004;3:55e60. [11] Taussky D, Austen L, Toi A, et al. Sequential evaluation of prostate edema after permanent seed prostate brachytherapy using CT-MRI fusion. Int J Radiat Oncol Biol Phys 2005;62:974e980. [12] Pinkawa M, Asadpour B, Gagel B, et al. Evaluation of source displacement and doseevolume changes after permanent prostate brachytherapy with stranded seeds. Radiother Oncol 2007;84: 190e196. [13] Ali I, Algan O, Thompson S, et al. A comparative study of seed localization and dose calculation on pre and post-implantation ultrasound and CT images for low-dose-rate prostate brachytherapy. Phys Med Biol 2009;54:5595e5611. [14] Papagikos MA, Deguzman AF, Rossi PJ, et al. Dosimetric quantifiers for low-dose-rate prostate brachytherapy: Is V100 superior to D90? Brachytherapy 2005;4:252e258. [15] Miles EF, Nelson JW, Alkaissi AK, et al. Equivalent uniform dose, D90, and V100 correlation with biochemical control after lowdose-rate prostate brachytherapy for clinically low-risk prostate cancer. Brachytherapy 2008;7:206e211. [16] Buskirk SJ, Pinkstaff DM, Petrou SP, et al. Acute urinary retention after transperineal template guided prostate biopsy. Int J Radiat Oncol Biol Phys 2004;59:1360e1366. [17] Stone NM, Stock RG. Complications following permanent prostate brachytherapy. Eur Urol 2002;41:427e433. [18] Eapen L, Kayser C, Deshaies Y, et al. Correlating the degree of needle trauma during prostate brachytherapy and the development of urinary toxicity. Int J Radiat Oncol Biol Phys 2004;59:1392e1394. [19] Lee N, Wuu CS, Brody R, et al. Factors predicting for postimplantation urinary retention after permanent prostate brachytherapy. Int J Radiat Oncol Biol Phys 2000;48:1457e1460.
Conclusion The visibility of the 9011 seeds and/or needles on ultrasound, fluoroscopy, and CT images is clinically acceptable. Use of the narrower seeds is likely to improve CT post-plan dosimetry by reducing artifacts and aiding the identification of closely spaced seeds. Further study is required into the suitability of MR images alone for post-plan dosimetry using the 9011 model seeds.
Acknowledgments The authors acknowledge radiologists Dr. Jonathan Smith and Dr. Brendan Carey for their assistance in the assessment of the clinical acceptability of the fluoroscopy images acquired for this work.
G. Roberts et al. / Brachytherapy 11 (2012) 460e467 [20] Bucci J, Morris WJ, Keyes M, et al. Predictive factors of urinary retention following prostate brachytherapy. Int J Radiat Oncol Biol Phys 2002;53:91e98. [21] Neill M, Studer G, Le L, et al. The nature and extent of urinary morbidity in relation to prostate brachytherapy urethral dosimetry. Brachytherapy 2006;6:173e179. [22] Wong J, Sylvester JE, Grimm PD, et al. First report on the use of a thinner I-125 radioactive seed within 20 gauge needles for permanent seed prostate brachytherapy: Evaluation of post-implant dosimetry and acute toxicity. Int J Radiat Oncol Biol Phys 2009; 75:S325. [23] Heintz BH, Wallace RE, Hevezi JM. Comparison of I-125 sources used for permanent interstitial implant. Med Phys 2001;28:671e682. [24] Meigooni AS, Luerman CM, Sowards KT. Evaluation of the dose distribution for prostate implants using various I-125 and Pd-103 sources. Med Phys 2009;36:2267e2277.
467
[25] Al-Qaisieh B, Smith DW, Brearley E, et al. Comprehensive I-125 multi-seed comparison for prostate brachytherapy: Dosimetry and visibility analysis. Radiother Oncol 2007;84:140e147. [26] Siebert FA, De Brabandere M, Kirisits C, et al. Phantom investigations on CT seed imaging for interstitial brachytherapy. Radiother Oncol 2007;85:316e323. [27] Rickey DW, Picot PA, Christopher D, et al. A wall-less vessel phantom for Doppler ultrasound studies. Ultrasound Med Biol 1995;21: 1163e1176. [28] Haacke EM, Brown RW, Thompson MR, et al. Magnetic field inhomogeneity effects and T2 dephasing. In: Haacke EM, editor. Magnetic resonance imaging: physical principles and sequence design. New York, NY: Wiley-Liss; 1999. p. 570e580. [29] De Brabandere M, Kirisitis C, Peeters R, et al. Accuracy of seed reconstruction in prostate postplanning studied with a CT- and MRI-compatible phantom. Radiother Oncol 2006;79:190e197.