Merrick, G. S.; Butler, W. M.; Dorsey, A. T.; Walbert, H. L.. Prostatic conformal brachytherapy: I-125/Pd-103 postopera- tive dosimetric analysis. Radiat. Oncol.
Int. J. Radiation Oncology Biol. Phys., Vol. 41, No. 1, pp. 217–222, 1998 Copyright © 1998 Elsevier Science Inc. Printed in the USA. All rights reserved 0360-3016/98 $19.00 1 .00
PII S0360-3016(97)00951-6
●
Physics Contribution QUALITY ASSURANCE CALIBRATION OF 125I RAPID STRAND IN A STERILE ENVIRONMENT WAYNE M. BUTLER, PH.D.,* ANTHONY T. DORSEY, M.S.,* KEITH R. NELSON† GREGORY S. MERRICK, M.D.*‡
AND
*Schiffler Oncology Center, Wheeling Hospital, Wheeling, WV; †Standard Imaging, Inc., Middleton, WI; ‡Division of Radiation Oncology and Biophysics, George Washington University Medical Center, Washington, DC Purpose: 125I seeds encased in stiffened absorbable suture material, commercially available from Amersham Healthcare as 125I Rapid Strand, are not readily calibrated because of the necessity of maintaining the sterility and integrity of the Rapid Strand. A method is proposed to verify the activity of 125I seeds in Rapid Strand in a sterile environment and to provide quality assurance of the resultant loading by autoradiographing loaded needles. Methods and Materials: A sterilizable insert for a Standard Imaging HDR 1000 Plus well chamber was designed to accommodate Amersham’s plastic spacing jig that holds the strand. The insert has a cylindrical lead sleeve allowing five of the 10 seeds in the jig to be exposed within the well chamber. A grooved tray for holding a batch of 30 needles was designed for autoradiographing the implant set. Results: Position-dependent calibration factors for each of the seed wells in the jig were determined; then these individual factors were combined into a single chamber factor of 21.6 pA/mCi. Starting from the most distal spacing jig slot, relative position factors were 0.99, 1.00, 0.99, 0.93, 0.73, and 0.13 for the six positions which produced a nonnegligible signal. Anisotropy in the chamber factor was determined by rotating the well in 30° increments about the seeds in the jig. The chamber factor showed only a 0.2% variation with rotational angle. Attenuation due to the Vicryl suture cladding material was 0.2% and was measured by stripping the Vicryl from the strand and remeasuring the chamber factor with the seeds in their original orientation. In the operating room, charge was collected from sterile Rapid Strands for a fixed time between 30 and 50 s and the measured ion chamber current was compared with the value predicted from Amersham’s nominal activity. The average deviation between nominal and measured activity of 73 Rapid Strands tested was 10.5% 6 2.2%. For single measurements, the maximum and minimum deviations observed were 14.8% and 23.7%, respectively. Autoradiographs taken of the entire implant set on an aluminum tray milled to hold the needles confirmed the actual seed loadings. Conclusions: The Rapid Strand calibration procedure described maintains the sterility and integrity of 125I Rapid Strands and verifies that the manufacturer’s stated activity is accurate to within 5%. © 1998 Elsevier Science Inc. 125
I Rapid Strand, Brachytherapy source calibration, Autoradiograph, Quality assurance.
INTRODUCTION The advent of transperineal ultrasound-guided conformal brachytherapy techniques combined with sophisticated computer treatment planning has resulted in a resurgence of interest in this modality of therapy for early-stage carcinoma of the prostate gland. Measured in terms of local control and freedom from biochemical failure, the results of conformal brachytherapy have been found to be at least as favorable as the most positive radical prostatectomy series (1– 4). In most procedures, prostate brachytherapy is performed with loose 125I or 103Pd seeds. In early 1995, Amersham Healthcare introduced radioactive 125I seeds embedded with 1-cm spacing in a stiffened Vicryl suture material (polyglactin). After stiffening and sterilization with ethylene oxide, 10 seed lengths of this material, called Rapid
Strand, are housed in a plastic spacing jig and shipped within a stainless-steel tube. This technologic advancement of attaching the radioactive seeds to an absorbable suture not only decreases the preparation time for the implant, but, more importantly, results in less radionuclide migration in the prostate/periprostatic region. For sources purchased in a sterile configuration, the AAPM Task Group 40 recommendation (5) is that a single nonsterile seed from each designated strength grouping used to manufacture the sterile sources be purchased and calibrated by the user. This is a concession to the difficulty of calibrating sterile source trains such as Rapid Strand, but this concession violates the spirit of the overall recommendation that a random sample of at least 10% of radioactive seeds be verified prior to implanta-
Reprint requests to: Wayne M. Butler, Ph.D., Schiffler Oncology Center, Wheeling Hospital, Medical Park, Wheeling, WV
26003. Accepted for publication 10 November 1997. 217
218
I. J. Radiation Oncology
●
Biology
●
Physics
tion. Amersham calibrates individual seeds, sorts them into batches within a designated strength range, and assembles a 10-seed Rapid Strand from seeds within a given range. These seeds have the nominal activity of the batch average, but a calibrated activity of the intact Rapid Strand is not available from the manufacturer. The source strength or activity of brachytherapy seeds containing radionuclides which emit low-energy photons may be verified using a well ionization chamber with direct or secondary traceability to national standards in the case of 125 I, or remote traceability by establishing a local standard using the seed manufacturer’s calibrated seeds in the case of 103 Pd. Loose seeds are shipped in a nonsterile package but may be autoclaved at any time after user calibration, usually just prior to an implant procedure. However, such confirmation of the desired seed activity has been difficult to implement clinically when this level of quality assurance requires the sterility of the Rapid Strands to be compromised. In our experience, both excessive handling and resterilization of a strand tend to diminish the stiffness of the Vicryl suture material between the seeds. Furthermore, autoclaving introduces dimensional changes through water absorption at high temperature which are not fully reversed by cooling to room temperature. Either weak spots between seeds or swollen braided suture material over a seed can contribute to the Rapid Strand jamming within the needle when the strand is extruded into the patient. Even straight from the manufacturer’s sterile package and with minimal handling, considerable difficulty was initially reported as a result of jamming of strands within the implant needles. Refinements of technique reported by both Butler and Friedland in loading the Rapid Strand segments into the needles and in the timing and manipulation of the needle within the patient resolved the needle jamming issues (6, 7). One solution to the problem of calibrating the source strength of Rapid Strand seeds is to load them into the implant needles, slip a sterile plastic sleeve over the needle, measure the output from the needle in a well ionization chamber, then dispose of the sleeve after the measurement. The steel needle, however, attenuates the already weak signal by about two thirds, and uncertainty in the position of seeds within the needle introduces uncertainty in the measured output owing to linear anisotropy along the cylindrical axis of typical well chambers. The solution we have developed is to use a specially designed sterilizable insert for a well chamber that holds an intact Rapid Strand while still in the manufacturer’s spacing jig. We further supplement this strand calibration with a semiquantitative batch evaluation by autoradiographing all implant needles, loose seed or Rapid Strand, after they are loaded.
Volume 41, Number 1, 1998
Fig. 1. Well chamber insert with Rapid Strand spacing jig in place. The insert holds the Rapid Strand spacing jig so that the 10 125I seeds are on the axis of the chamber. When the jig is pushed all the way in, the lead sleeve shields exactly half the strand (between seed positions 5 and 6). The spacing jig is semicircular in cross section, and the hole through the insert is similarly shaped to prevent jig rotation.
chosen because its small size and lightweight shielding make it easy to transport to the operating room along with a portable electrometer. Only 12 cm deep, the well chamber is sufficiently shallow so that a portion of the 14-cmlong Rapid Strand spacing jig will protrude from the well for expedited insertion and retrieval. Symmetrically arrayed on the jig in spacing wells, seed centers are located 2.5 cm inboard from each end of the jig. The seeds themselves are 4.5 mm long. Since many measurements
METHODS AND MATERIALS A custom insert shown in Figs. 1 and 2 designed to fit into a Standard Imaging HDR 1000 Plus well chamber was used for all measurements. This well chamber was
Fig. 2. The well chamber insert relative to the well chamber.
Rapid strand QA
● W. M. BUTLER et al.
Table 1. Well chamber factors for
219
125
I seeds
Slot number
1
2
3
4
5
6
7
Total
pA/mCi Normalized
4.48 0.99
4.52 1
4.47 0.99
4.21 0.93
3.29 0.73
0.6 0.13
0.06 0.01
21.63 4.78
were made with loose seeds resting in spacing wells of the jig, all measurements were made with the well chamber lying on its side, resting on a cradle cut from Styrofoam. The shallow depth of the chamber prevents a complete Rapid Strand from fitting within the sensitive volume of the well chamber. By using a lead and aluminum sleeve positioned to cut off the radiation between seed positions 5 and 6, nominally half the Rapid Strand may be calibrated at one time. The aluminum sleeve and shoulder on the distal end of the insert provide shielding for the operator from the last seed protruding outside the chamber. Since the Amersham spacing jigs are semicircular in cross section, plastic tabs were glued at the ends of the shielding sleeve to prevent rotation of the jig and to keep the seeds aligned along the axis of the well chamber. The well chamber was connected to a calibrated Victoreen 525 electrometer. Although ion current is the most important quantity to be determined, the long settling time for the low currents expected (,10 pA) and resolution of only 10 fA led us perform all measurements by integrating charge collected for a period of time measured by a stopwatch. RESULTS Chamber factors Well chambers show an axial positional sensitivity in signal output for a given source (8). The spacing jig will hold the seeds in a fixed position and allow a separate 125I calibration factor to be determined for each position. An overall chamber factor is then calculated as the sum of the individual positional factors. To determine the positional dependence of each possible seed location in the well chamber, individual 125I seeds calibrated by Amersham Healthcare in the activity range 0.67– 0.82 mCi were placed in successive seed wells of an empty spacing jig. Each charge reading in nC was corrected for chamber temperature and pressure and then converted to specific ion current in pA/mCi. Calibrated seed strengths were appropriately decayed to the time of the measurement. During the manufacture of 125I seeds, radioactive iodine is adsorbed onto silver wire, but the adsorption is not always uniform. To minimize the effect of this asymmetric adsorption, which will produce a linear and rotational radiation anisotropy around a seed, each seed was placed in a variety of orientations within the seed wells, and this procedure was repeated using four calibrated seeds. Individual seed well calibration factors and the overall
chamber factor in pA/mCi of individual seed activity for an 125I Rapid Strand on its spacing jig are listed in Table 1. These calibration factors are also listed normalized to the maximum seed well factor. Position factors for seed wells numbered beyond 7 were negligible because of the lead sleeve on the insert, which largely limits the measured signal to the first five positions. Well chamber isotropy To determine the optimal geometry of the insert containing a Rapid Strand within the well chamber, the well chamber was arbitrarily marked in 30° increments, and the signal from intact Rapid Strands with nominal individual seed activity of 0.67 mCi was measured with the insert held stationary and the chamber rotated to each angular position. The relative chamber output is isotropic to rotation around the jig and insert to 1.000 6 0.002. Although the chamber is isotropic to rotation about the jig, all measurements and calibrations were made in a single fixed orientation. Attenuation due to braided vicryl seed cladding The braided Vicryl suture material cladding the 125I seeds should slightly attenuate the low-kilovoltage photon emissions. The amount of attenuation was determined by first measuring the well chamber output from seeds in intact Rapid Strands. Next, using forceps, the seeds were pushed out from the braided Vicryl material one by one and the seeds were replaced in the spacing jig in their original slots and with the same lengthwise orientation. The ratio of the output from Vicryl-clad Rapid Strand seeds to the stripped seeds gave an average attenuation factor of 0.998 6 0.005. This effect is negligible and was not considered in clinical measurements of Rapid Strand activity. System linearity The procedure for determining the chamber factor for each seed slot in the jig involved measuring the charge over an extended period of time (5–30 min/trial). However, the ultimate goal was to use the overall chamber factor for measurements of Rapid Strands in the operating room using exposures of ,1 min duration. Therefore, a system linearity check was performed by collecting charge from Rapid
Table 2. Linearity of calculated ion current vs. charge integration time Charge integration time (s) Normalized current
4000 1
400 0.998
40 0.988
220
I. J. Radiation Oncology
●
Biology
●
Physics
Volume 41, Number 1, 1998
Table 4. % deviation between measured and predicted ion currents
Table 3. Predicted vs. measured current Seed Well No. 1 2 3 4 Total
Seed Activity (mCi)
Slot Factor (pA/mCi)
Predicted Current (pA)
0.468 0.773 0.477 0.785
4.479 4.524 4.472 4.206
2.095 3.499 2.133 3.302 11.03
Measured Current (pA)
11.09
Strands for different time intervals and comparing the calculated ion currents. The results, normalized to the ion current measured for the longest of the three time intervals tested, are shown in Table 2. The ion currents, which range from 5 to 8 pA for seed activities in the normal clinical range of 0.25– 0.35 mCi for prostate brachytherapy, are underestimated by about 1% when integrating for the short time interval used in the operating room. Algorithm check To demonstrate that individually determined seed slot calibration factors in pA/mCi can be combined to give an overall chamber factor, the ion current due to four calibrated seeds in various spacing wells along the spacing jig was determined. For each combination of seed and slot, the product of seed activity and slot factor gives a predicted ion current from each occupied location. The sum of these for all occupied locations should equal the total chamber current. A sample calculation for one seed arrangement compared with the value measured for that arrangement is listed in Table 3. The difference between measured and predicted values may be attributed to seed anisotropies in the measured configuration relative to average values used in determining the slot factors. Clinical quality assurance tests In the operating room, the previously sterilized well chamber insert is removed from its package and placed into the well chamber using a sterile technique. The nominal activity of the Rapid Strand seeds is decayed to the time of measurement, and a predicted electrometer reading is calculated using the following equation: R 5 A z CF t TPF z EF where R 5 the electrometer charge reading in pC; t 5 the measurement time in seconds; A 5 the decayed seed activity in mCi; CF 5 the Rapid Strand chamber factor in pA/mCi; TPF 5 the temperature–pressure correction factor; and EF 5 the electrometer correction factor, which includes the linearity adjustment discussed above. This predicted value for the ion current may then be compared with measured ion currents from Rapid Strands being prepared for patient use. For each Rapid Strand tested, only a single measurement was made, and a charge . 200 pC could be
No. of Strands
Maximum Dev. (%)
Minimum Dev. (%)
Average Dev. (%)
SD of Average (%)
73
14.8
23.8
10.50
2.2
collected in , 45 s. Although most of our clinical measurements were performed by one person using a single sterile hand to manipulate Rapid Strands in and out of the chamber and the other hand to operate the electrometer and record data, two people are more efficient at performing the task. A summary of our results comparing measured ion currents and those predicted from Amersham’s stated batch activity is presented in Table 4. The range of activities used in these measurements was 0.22– 0.32 mCi/seed, and these produce currents from 5 to 7 pA. The measured average conforms to the nominal manufacturer’s activity within 0.5 6 2.2%. The magnitude of the standard deviation is due primarily to the 4% spread in activities about the mean in the manufacturer’s batches. Also noteworthy is the fact that every individual one-time measurement was within 5% of the predicted value (range 4.8% high to 3.8% low) and that the measured average of Rapid Strand source strength used for any individual patient was within 2.5% of the calculated value used in preplanning. Autoradiographic confirmation of needle loading After all the implant needles are loaded with either free seeds or Rapid Strand segments, their loading is confirmed by autoradiographing the needles using 8 3 10-inch Kodak AR film. We designed a simple device to batch-autoradiograph up to 30 preloaded needles. The needles are placed in grooves 2 mm wide and 2 mm deep milled across the width of an 8.5 3 10-inch (203 3 254-mm) aluminum plate that is 13 mm thick. This plate is presterilized, and once the needles are in place, the plate and needle array are covered with a sheet of sterile plastic or paper. An unmilled aluminum cover plate of the same size, but with a handle attached to the back, has a sheet of the Ready-Pack AR film taped to the front. The film and cover plate are placed over the sterility-protected needle array and held in place for about 15–20 s. While the film is being developed, the needles are inserted into their assigned grid positions in the storage box. Although the contents of a needle may be checked for the proper number of seeds and spacers by measuring the distance from the hub to the stylet, this check does not guard against improper sequencing in needles loaded with loose seeds. Also, owing to the variability in the length of the plug placed at the tip of the needle, physical measurements may fail to detect mistakes in the number of seeds or spacers used. One such film of 29 needles and 144 seeds (as part of a large implant involving 36 needles and 168 seeds) is shown on the accompanying autoradiograph in Fig. 3. At the top of the film, six needles containing six free seeds each have additional spacers near the middle of the seed train. In
Rapid strand QA
Fig. 3. Autoradiograph of 29 needles loaded with 125I seeds. The first six needles from the top intentionally deviate from alternating free seeds and spacers and reduce the linear density of seeds by placing an extra spacer around the third or fourth seed from the end. All but six of the remaining needles contain Rapid Strand sequences.
this case, the loading is intentionally used for needles inserted near the centerline of large prostate glands to keep the midgland dose under 150% of the prescribed minimal peripheral dose (9). Through 2.5 years experience with Rapid Strands, we have failed to identify any misloaded strands, but we have identified errors in loading needles with loose seeds at the rate of 1/200 needles. Almost all errors (.90%) have involved missing spacers, but errors in sequencing and seed count have also been noted. All loading errors are easily corrected by expelling the source train and reloading the needle. The annotated autoradiograph also provides a qualitative verification of seed activity and is placed in the patient’s chart as a quality assurance document. DISCUSSION Introduction in early 1995 of 125I seeds embedded in a rigid braided Vicryl suture represented a technologic advancement by decreasing the migration of seeds in the capsular/periprostatic region and embolization to distant sites within the body. Rapid Strand–loaded needles are not recommended for use along the midgland region, because of
● W. M. BUTLER et al.
221
concern for strands traversing the urethra with resultant irritative symptomatology. Furthermore, loose seeds within the corpus of the gland appear to hold their position well over time, based on the results of an ongoing study at this institution. The use of Rapid Strands has been shown to result in more accurate placement of seeds, especially in the extracapsular/periprostatic regions (10). To date, there has been no head-to-head comparison of postoperative dosimetry comparing loose 125I seeds and Rapid Strand. Merrick et al. (9) reported postoperative dosimetric results for 125I implants. With 125I Rapid Strands comprising 65% of the implanted seeds, 96% of the prostate volume (range 80 – 99%) received the full minimal peripheral dose (mPd). Willins and Wallner (11) reported an average of 84%, while Moerland et al. (12) reported 60% of the prostate receiving the prescribed dose. Both of these investigations used loose 125 I seeds. In our opinion, the utility of Rapid Strand is the fact that seeds can be confidently placed along the capsule and in the periprostatic region without concern for migration or embolization to distant sites. Depositing seeds accurately at the planned locations is an absolute requirement for good postoperative dosimetry. That these seeds stay in their planned locations is one very important factor in achieving high-quality postoperative dosimetric results. Task Group 40 of the AAPM has recommended that at least 10% of permanently implanted loose seeds be verified according to source strength. Here, we have described a technique for extending this recommendation to the verification of 125I source strength in sterile Rapid Strands. Our method is to use a specially designed sterilizable insert for a well chamber that holds an intact Rapid Strand on its spacing jig. The insert effectively truncates the measurement at half of the strand. This technique allows for accurate calibrations without compromising the sterility of the strand. Another calibration technique has been reported by Feygelman et al. (13). Our Rapid Strand calibration procedure using individually calibrated seeds has been shown to be consistent with the manufacturer’s nominal activity of seeds within the strand to within 65% at the 95% confidence level. Because of this demonstrated consistency, it would be reasonable to determine a well chamber calibration factor in terms of a local standard by assuming that the nominal activity of a Rapid Strand is exact. The accuracy of the chamber factor so determined could then be refined through output measurements of Rapid Strands from succeeding batches. Autoradiographing brachytherapy source trains for quality assurance has been used with other isotopes (14), but autoradiographs of seeds intended for permanent implants have been limited to batch assays of loose seeds prior to loading (15). Not only does autoradiographing batches of preloaded needles prior to an implant provide a crude activity assay, but it has also been found to be useful in verifying seed loading and in finding and correcting loading errors. No errors were found in Rapid Strand–loaded needles, but since Rapid Strand should not be used in areas of the prostate where there is a likelihood of traversing the
222
I. J. Radiation Oncology
●
Biology
●
Physics
urethra, implants that attempt to follow the modified uniform loading approach pioneered by the Seattle group (16)
Volume 41, Number 1, 1998
will use some needles loaded with free seeds and spacers near the centerline of the gland.
REFERENCES 1. Blasko, J. C.; Wallner, K.; Grimm, P. D.; Ragde, H. Prostate specific antigen based disease control following ultrasound guided iodine-125 implantation for Stage T1/T2 prostatic carcinoma. J. Urol. 154:1096 –1099; 1995. 2. Wallner, K.; Roy, J.; Zelefsky, M.; Fuks, Z.; Harrison, L. Short-term freedom from disease progression after I-125 prostate implantation. Int. J. Radiat. Oncol. Biol. Phys. 30:405– 409; 1994. 3. Wallner, K.; Roy, J.; Harrison, L. Tumor control and morbidity following transperineal I-125 implantation for Stage T1/T2 prostatic carcinoma. J. Clin. Oncol. 14:449 – 453; 1996. 4. Dattoli, M.; Wallner, K.; Sorace, R.; Koval, J.; Cash, J.; Acosta, R.; Brown, C.; Etheridge, J.; Binder, M.; Brunell, E. R.; Kirwan, N.; Sanches, S.; Stein, D.; Wasserman, S. Pd-103 brachytherapy and external beam radiation for clinically localized high-risk prostatic carcinoma. Int. J. Radiat. Oncol. Biol. Phys. 35:875– 879; 1996. 5. Kutcher, G. J.; Coia, L.; Gillin, M.; Hanson, W. F.; Leibel, S.; Morton, R. J.; Palta, J. R.; Purdy, J. A.; Reinstein, L. E.; Svensson, G. K.; Weller, M.; Wingfield, L. Comprehensive QA for radiation oncology: Report of AAPM Radiation Therapy Committee Task Group 40. Med. Phys. 21:581– 618; 1994. 6. Butler, W. M.; Merrick, G. S. I-125 Rapid Strand loading technique. Radiat. Oncol. Invest. 4:48 – 49; 1996. 7. Friedland, J. L.; Feygelman, V.; Haller, E. M.; Bradley, L. M.; Sanders, R. M.; Noriega, B. K. Problems with rigid seed strand lodging during prostate implantation: A proposed mechanism and solution. Med. Dosim. 22:17–21; 1997. 8. DeWerd, L. A.; Thomadsen, B. R. Source strength standards and calibration of HDR/PDR sources. In: Williamson, J. F.;
9.
10.
11.
12.
13.
14.
15.
16.
Thomadsen, B. R.; Nath, R., eds. Brachytherapy physics. Madison, WI: Medical Physics; 1995:541–555. Merrick, G. S.; Butler, W. M.; Dorsey, A. T.; Walbert, H. L. Prostatic conformal brachytherapy: I-125/Pd-103 postoperative dosimetric analysis. Radiat. Oncol. Invest. (In press). Roberson, P. L.; Narayana, V.; McShan, D. L.; Winfield, R. J.; McLaughlin, P. W. Source placement error for permanent implant of the prostate. Med. Phys. 24:251–257; 1997. Willins, K.; Wallner, K. CT based dosimetry for transperineal I-125 prostatic brachytherapy. Int. J. Radiat. Oncol. Biol. Phys. 39:347–353; 1997. Moerland, M. A.; Wijrdeman, H. K.; Beersma, R.; Bakker, C. J.; Battermann, J. J. Evaluation of permanent I-125 prostate implants using radiography and magnetic resonance imaging. Int. J. Radiat. Oncol. Biol. Phys. 37:927–993; 1997. Feygelman, V.; Noriega, B. K.; Sanders, R. M.; Friedland, J. W. A simple method for verifying activity of iodine-125 seeds in rigid absorbable suture. Med. Dosim. 21:261–262; 1996. Beach, J. L.; Maruyama, Y. A simple method using film monitoring for Cf-252 brachytherapy: Technical note. Radiat. Med. 2:140 –142; 1984. Berkley, L. W. Design and treatment planning for permanent implants. In: Williamson, J. F.; Thomadsen, B. R.; Nath, R., eds. Brachytherapy physics. Madison, WI: Medical Physics; 1995:427– 439. Grimm, P. D.; Blasko, J. C.; Ragde, H. Ultrasound guided transperineal implantation of iodine-125 and palladium-103 for the treatment of early stage prostate cancer. Atlas Urol. Clin. North Am. 2:113–125; 1994.
