Reducing seed waste and increasing value of dynamic intraoperative implantation of Pd‐103 seeds in prostate brachytherapy

Abstract Several nomograms exist for ordering palladium‐103 seeds for permanent prostate seed implants (PSI). Excess seeds from PSIs pose additional radiation safety risks and increase the cost of care. This study compared five nomograms to clinical data from dynamic modified‐peripheral intraoperative PSI to determine (a) the cause of excess seeds and (b) the optimal nomogram for our institution. Pre‐ and intraoperative patient data were collected for monotherapy PSIs and compiled into a clinical database. All patients were prescribed 125 Gy with dose coverage of D90% = 100% to the planning target volume (PTV) using 103Pd seeds with mean air‐kerma strength (SK¯) of 2 U. Seeds were ordered based upon an in‐house nomogram as a function of preoperative prostate volume and prescription dose. Preoperative prostate volume was assessed with transrectal ultrasound. If any of the following four conditions were not met: (a) preoperative volume = intraoperative volume, (b) D90% = 100%, (c) SK¯=2U, and (d) seed ordering matched the in‐house nomogram, then a normalization factor was applied to the number of seeds used intraoperatively to meet all four conditions. Four published nomograms, an in‐house nomogram, and the normalized number of implanted seeds for each patient were plotted against intraoperative prostate volume. Of the 226 patients, 223 had excess seeds at the completion of their PSI. On average, 25.7 ± 9.9% of ordered seeds were not implanted. Excess seeds were separated into two categories, accounted‐for excess, determined by the four normalization factors, and residual excess, assumed to be due to overordering. The upper 99.9% CI linear fit of the normalized clinical data plus a 5% “cushion” may provide a more reasonable nomogram for 103Pd seed ordering for our institution. Nomograms customized for individual institutions may reduce seed waste, thereby reducing radiation safety risks and increasing the value of prostate brachytherapy.


| INTRODUCTION
Many treatment options, such as radical prostatectomy, external beam radiation therapy (EBRT), brachytherapy, hormonal treatment, and active surveillance, exist for patients with prostate cancer. 1 Transperineal interstitial permanent prostate brachytherapy (TIPPB) for early stage prostate cancer is an outpatient procedure involving insertion of radioactive seeds into the prostate under transrectal ultrasound guidance. 1 Current radioisotopes used in TIPPB include 125 I, 103 Pd, and 131 Cs. [1][2][3][4] At our institution, 103 Pd seeds are implanted using a dynamic intraoperative technique 1-3 and the workflow for seed ordering, implantation, and postimplant dosimetry is similar to that described in AAPM TG-64. 2 The workflow is as follows: (a) 103  Preoperative imaging of the patient's prostate is used to determine the number of 103 Pd seeds to be ordered. At our institution, a nomogram comparing total recommended air-kerma strength (U) and prostate volume (cm 3 ) is used to determine the recommended total air-kerma strength for the implant. The number of required seeds is found by dividing the total air-kerma strength by the requested air-kerma strength per seed. In our clinical experience, we have often found that there are a substantial number of excess 103 Pd seeds after prostate seed implants. Reducing the number of unused seeds can reduce radiation safety risks, reduce cost, and increase the value of prostate brachytherapy in the context of value-based medicine, defined as, "the practice of medicine incorporating the highest level of evidence-based data with the patient-perceived value conferred by healthcare interventions for the resources expended." 5 The purpose of this study was, (a) determine the cause of excess 103 Pd seeds in dynamic intraoperative TIPPB and (b) determine the optimal nomogram for our institution.

| MATERIALS AND METHODS
Prostate cancer patients who received monotherapy using 103 Pd to a dose of 125 Gy were retrospectively included in the chart review.
Relevant dosimetric and volumetric pre-, intra-, and postoperative measurements were compiled into a clinical database.

2.A | Excess seeds
The clinical database was used to determine the number of excess seeds for each patient. The number of excess seeds (N XS ) is defined in eq. (1).
N XS ¼ ðnumber of seeds orderedÞ À ðnumber of seeds used intra À operativelyÞ The total number of excess seeds can be separated into two categories, accounted-for excess (A XS ) and residual excess (R XS ). The quantity A XS is equal to the sum of the number of excess seeds remaining after an implant (n) which can be attributed to a specific reason (i) [eq. (2)]. The value of n i can be positive or negative.
Excess seeds that could not be attributed to a specific cause were considered "residual." The quantity R XS is equal to the number of excess seeds less the number of accounted-for excess seeds [eq. (3)].
We identified four potential causes of excess seeds which are summarized in Table 1 and described in greater detail below. The value n vol was calculated by taking the difference between the numbers of seeds which should have been ordered for each respective volume (preoperative and intraoperative) based upon the in-house nomogram. If preoperative prostate volume was greater than intraoperative prostate volume, additional, unnecessary seeds were ordered, i.e., n vol [ . If the intraoperative prostate volume was greater than the preoperative volume, more seeds were used than anticipated, which lowered the expected number of excess seeds at the end of the procedure, i.e., n vol .
The value n order was the difference between what was ordered and what should have been ordered based on the in-house nomogram and the preoperative prostate volume. The output of the nomogram is total recommended air-kerma strength. When this value is converted into number of seeds using vendor specified air-kerma strength per seed, it is not necessarily an integer. For simplicity in ordering and to insure enough seeds were present for the procedure, nomogram results were frequently rounded up to the nearest multiple of five. If more seeds were ordered than indicated by the nomogram, the result is a known excess compared to the nomogram, i.e., n order . If fewer seeds were ordered than indicated by the nomogram, the result is a known seed deficit, i.e., n order .
T A B L E 1 Variables which contribute to deviations between the expected number of seeds used intraoperatively and actual number of seeds used intraoperatively.

Reasons for excess seeds Description
Number of excess seeds due to reason (i)

Change in prostate volume
Preoperative prostate volume ≠ intraoperative prostate volume n vol

Mean air-kerma strength
Mean air-kerma strength (S K ) ≠ 2 U per seed n U

Ordering
Number of ordered seeds deviated from nomogram n order All patients were prescribed 125 Gy with dose coverage of D90% = 100% to the planning target volume (PTV). If the intraoperative D90% was less than 100%, i.e., a "cold implant," then not enough seeds were used during the implant resulting in a known seed excess at the end of the procedure, i.e., n D90% . If the intraoperative D90% was greater than 100%, more seeds were used than anticipated which lowered the number of excess seeds at the end of the procedure, i.e., n D90% . If the mean air-kerma strength (S K ) for a batch of seeds was greater than 2 U per seed, fewer seeds were required to meet the prescription dose of D90% = 100% to the PTV resulting in a known seed excess, i.e., n U . If the mean air-kerma strength (S K ) for a batch of seeds was less than 2 U per seed, a greater number of seeds were required to meet the prescription dose of D90% = 100% to the PTV resulting in a known seed deficit, i.e., n U .

2.B | Clinical data normalization parameters
Once the cause of excess seeds was determined, the number of seeds used intraoperatively per patient was linearly normalized such that S K ¼ 2 U and the intraoperative D90% was equal to 100%. The  Table 2 lists patient-specific assessments for three sample patients. Table 3 shows the calculation performed using the assessments in Table 2 to determine n i and subsequently N XS , A XS , and R XS . A negative sign in Table 3 indicates that variable contributed positively toward R XS , that is, there should have been more seeds at the end of the procedure due to that specific variable.

3.A | Excess seeds
Of the 226 prostate seed implants, 98.6% (n = 223) had excess seeds. On average (±1 standard deviation), there were 29.2 ± 13.2 excess seeds after the completion of a prostate seed implant. On average, 25.7 ± 9.9% of ordered seeds were not implanted. The percentage of ordered seeds which were wasted is the quotient of the number of excess seeds to the number of seeds ordered (eq. 4). Figure 1 shows the distribution of the percentage of ordered seeds which were wasted.
Percentage of ordered seeds which were wasted = # of excess seeds # of seeds ordered Â 100% We found that none of the four parameters identified above contributed substantially to seed excess. In the cases where the intraoperative volume was greater than preoperative volume (53% of implants, n = 117), more seeds were used than anticipated to accommodate the larger prostate. In the cases where the implant had a D90% > 100%, i.e., a "hot implant" (72% of implants, n = 159), more seeds were used than anticipated which should have contributed towards the final number of excess seeds. In the cases where the mean air-kerma strength was greater than 2 U (43% of implants, n = 98), more seeds were used than anticipated which should have contributed towards the number of excess seeds. The value of n i was on average negative for each "i." The consequence of n i \0 means the number of residual-excess seeds was on average greater than the physical number of excess seeds. On average, 30.7 ± 7.1% of ordered seeds were R XS . Table 4 shows the average N XS , R XS , and n i .

AUTHORSHIP
All persons who meet authorship criteria are listed as authors, and all authors certify that they have participated sufficiently in the work to take public responsibility for the content, including participation in the concept, design, analysis, writing, or revision of the manuscript.
Furthermore, each author certifies that this material or similar material has not been and will not be submitted to or published in any other publication.

CONFLI CTS OF INTEREST
The authors whose names are listed immediately below certify that they have NO affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants;