- Email: [email protected]
Prostate HDR brachytherapy catheter displacement between planning and treatment delivery
Radiotherapy and Oncology 101 (2011) 490–494
Contents lists available at SciVerse ScienceDirect
Radiotherapy and Oncology journal homepage: www.thegreenjournal.com
Prostate cancer brachytherapy
Prostate HDR brachytherapy catheter displacement between planning and treatment delivery May Whitaker a,⇑, George Hruby a,b, Aimee Lovett a, Nitya Patanjali a a
Department of Radiation Oncology, Sydney Cancer Centre; and b Division of Medicine, University of Sydney, NSW, Australia
a r t i c l e
i n f o
Article history: Received 27 October 2010 Received in revised form 3 August 2011 Accepted 5 August 2011 Available online 31 August 2011 Keywords: HDR prostate brachytherapy Catheter displacement Movement
a b s t r a c t Background and purpose: HDR brachytherapy is used as a conformal boost for treating prostate cancer. Given the large doses delivered, it is critical that the volume treated matches that planned. Our outpatient protocol comprises two 9 Gy fractions, two weeks apart. We prospectively assessed catheter displacement between CT planning and treatment delivery. Materials and methods: Three fiducial markers and the catheters were implanted under transrectal ultrasound guidance. Metal marker wires were inserted into 4 reference catheters before CT; marker positions relative to each other and to the marker wires were measured from the CT scout. Measurements were repeated immediately prior to treatment delivery using pelvic X-ray with marker wires in the same reference catheters. Measurements from CT scout and film were compared. For displacements of 5 mm or more, indexer positions were adjusted prior to treatment delivery. Results: Results are based on 48 implants, in 25 patients. Median time from planning CT to treatment delivery was 254 min (range 81–367 min). Median catheter displacement was 7.5 mm (range 2.9– 23.9 mm), 67% of implants had displacement of 5 mm or greater. Displacements were predominantly caudal. Conclusions: Catheter displacement can occur in the 1–3 h between the planning CT scan and treatment. It is recommended that departments performing HDR prostate brachytherapy verify catheter positions immediately prior to treatment delivery. Crown Copyright Ó 2011 Published by Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology 101 (2011) 490–494
In high dose rate (HDR) brachytherapy the radiation dose decreases rapidly outside the treatment volume, making it an ideal technique for treating targets lying close to organs at risk such as the prostate gland. With imaging tools such as ultrasound and CT, together with a 3D treatment planning system and anatomical contouring, dose volume histograms are relatively quick and simple to generate. However, it is critical that the treated volume matches the planned volume due to the large doses delivered per fraction, otherwise there is a risk of over-treating critical organs such as the urethra or rectum, and under-treating the target. This has been demonstrated in work at Mt Vernon Cancer Centre in the UK [1] and Sir Charles Gairdner Hospital in Australia [2]. Clinically, caudal displacement of catheters may be responsible for the relatively high incidence of urethral strictures seen after HDR brachytherapy [3]. In this study, we prospectively assessed catheter movement between time of CT imaging and delivery of the first fraction. ⇑ Corresponding author. Address: Royal Prince Alfred Hospital, Department of Radiation Oncology, Building 27 Missenden Rd., Camperdown, NSW 2050, Australia. E-mail address: [email protected] (M. Whitaker).
Materials and methods Patient selection criteria Patient selection for HDR prostate brachytherapy boost included suitability for anaesthesia, life expectancy of at least 10 years, adequate baseline urinary function, and prostate volume less than 60 cc. Tumour characteristics included serum PSA of 10 lg/l or more; and/or Gleason Score of 7 or more; and no evidence of metastases see Table 1. Tumour treatment At the Sydney Cancer Centre (SCC), HDR brachytherapy is used in the management of prostate cancer as a conformal boost in conjunction with external beam radiation treatment (EBRT). The boost is delivered as two fractions of 9 Gy, two weeks apart, with EBRT of 46 Gy in 23 fractions starting approximately 4–5 days after the first brachytherapy treatment. A one day break in EBRT occurs on the day of the second brachytherapy implant and treatment. This regimen has been adopted for patient comfort, to avoid hospitalization, and to eliminate any inter-fraction implanted catheter movement.
0167-8140/$ - see front matter Crown Copyright Ó 2011 Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.radonc.2011.08.004
M. Whitaker et al. / Radiotherapy and Oncology 101 (2011) 490–494 Table 1 Tumour characteristics of the 25 patients involved in this study. PSA
<10 10–20 >20
8 15 2
Tumour stage
T1 T2 T3
7 17 1
Gleason score
6 3+4 4+3 8
1 10 13 1
No: 4
Yes: 21
Androgen deprivation
The implant procedures took place in our departmental treatment bunker which is operating-theatre equipped. All patients underwent either general or spinal anaesthesia and were placed in the lithotomy position. After insertion of an in-dwelling urinary catheter, 3 fiducial markers (CIVCO ACCULOCÒ, Orange City, USA) were placed transperineally into the prostate (2 at the base, 1 at mid or apex) under ultrasound guidance using a prostate stepper template (Nucletron, Veenendaal, Netherlands), at the first implant procedure only. Subsequently, the Pro-guide catheters (Nucletron, Veenendaal, Netherlands) were placed using the same template which was firmly sutured to the perineum at the end of the procedure (Fig. 1). The catheters were implanted to well beyond the base of the prostate (approximately 2 cm into the bladder) to allow for compensation of any displacement measured before treatment delivery. Routine cystoscopy was not performed, however, catheter incursion into the bladder was documented with sagittal ultrasound imaging. Once the catheters were labelled, and the distance from template to catheter hubs measured, the patient’s legs were removed from the stirrups and the patient was transferred (down a short corridor) to our CT scanner. The patient was positioned on the CT couch with his legs down, and the distance from template to catheter hubs was again measured and confirmed prior to scanning. The patient was then placed back on the hospital bed to await final transfer to the treatment room for dose delivery after treatment planning. Treatment planning was performed using the PLATO planning system (Nucletron, Veenendaal, Netherlands). The Radiation Oncologists (GH and NP) contoured the target and organs at risk, and the PLATO Inverse Planning Simulated Annealing algorithm was used to optimise the dwell positions and times.
Fig. 1. Catheters implanted in a patient using a prostate stepper template. The template is sutured to the patient’s skin and the template to catheter hub distance is measured.
491
Prior to treatment delivery, the distance from template to catheter hubs was measured a third and final time. An anterior– posterior (AP) digital X-ray image was taken with the patient’s legs down, and treatment was delivered in this position on the hospital bed. Each fraction was delivered several hours after implantation. The template was removed immediately following treatment, with the IDC being removed 2–3 h later following bladder irrigation. Equipment Equipment for all cases included the following – Microselectron Iridium-192 afterloader, PLATO planning software, Pro-guide catheters, Microtouch stepper-stabilizer unit (all Nucletron, Veenendaal, Netherlands) and B&K Ultrasound unit (B&K, Herlev, Denmark). The template to proximal catheter hub distance for 4 reference catheters was measured after each insertion with the legs still in the lithotomy position, to establish a baseline reading. This measurement was repeated with the legs down at the CT planning scan and again immediately prior to treatment. Planning was undertaken using a CT scan (Toshiba, Tustin, USA) with 3 mm slice thickness and the PLATO planning system. To compare catheter positions at the time of CT scanning and time of treatment delivery, metal marker wires were inserted into the same 4 reference catheters at CT and during the digital X-ray acquisition; seed positions relative to each other (measuring from the centre of each seed) and vertical measurement from the tip of each of the 4 marker wires were recorded from the CT scout image and the pre-treatment X-ray. Reference catheters were defined as the two most anterior/medial catheters and the two most lateral catheters. Seed to seed measurements as well as catheter to seed measurements were used to assess any implant displacement relative to the prostate. See Fig. 2. Measurement of displacement Internal seed and reference marker wire position measurements were repeated immediately prior to treatment delivery using an AP pelvic X-ray with the marker wires again inserted in the same 4 reference catheters used at the time of planning scan. Measurements from the CT scout and film were compared and the average value over the 4 reference catheters from each image was used as the displacement magnitude (Fig. 3).
Fig. 2. Measurement process for seed to seed and seed to catheter measurement. The distance from the tip of the catheter to the tip of all 3 fiducial markers is measured for the 4 catheters. These values are then averaged to produce a single value for the catheter to seed distance for the planning CT scout. The process is then repeated using the pre-treatment X-ray. The difference between the two values is recorded as the measured displacement.
492
Prostate brachytherapy
25.0
Measured shift distance (mm)
20.0
15.0
10.0
5.0
0.0
7 33
9
1 33
0 31
31
5
3 30
30
0
7 28
28
4
8 26
27
7 25
1
4 23
25
5
9 19
4 19
19
4
0 19
17
9
5 16
15
7
1 15
12
81
-5.0
Time elapsed between acquisition of CT scan and acquisition of digital x-ray (min) Fig. 3. The time elapsed between acquisition of the CT scout scan and the pre-treatment X-ray does not appear to be a consideration in the magnitude of catheter displacement.
It was assumed that catheter displacements occurred only in the longitudinal axis [2]. A radial shift of up to 20° will result in an error of approximately 1 mm in the displacement measurement. This was considered insignificant with respect to the method of displacement calculation. Orthogonal X-rays were initially attempted, however, accurate measurements could not be made due to the poor image quality; therefore, AP X-rays alone were used for displacement measurements. Shift protocol Based on the action threshold used at Mt. Vernon [4], displacements of 5 mm or more measured between planning CT and pretreatment X-ray were actioned. To compensate for any measured displacement greater than 5 mm, the source was digitally advanced further into the patient by altering the indexer length at the treatment console. Implanting the catheters past the prostate base into the bladder allowed for extra dwell positions beyond the prostate in the event of caudal shift. Physical re-insertion was not performed. Results The 25 patients in this study had a median age of 67 yr (range 47–78). Most men (n = 22) had intermediate risk prostate cancer. Three men had high risk disease, of whom two had 1 adverse feature each and one patient had 2 adverse features (T3 and PSA > 20). The median PSA was 12.4 (range 4–32). The tumour characteristics are detailed in Table 1. Our results are based on a total of 48 implants performed on 25 patients, with a median of 18 catheters (range 15–26) per implant. The median measured displacement was 7.5 mm (range 2.9– 23.9 mm), with shifts of greater than or equal to 5 mm occurring in 32 implants (67%). Displacement in the cephalad direction (as represented by a negative value) was uncommon (3 of 48 cases only) and was always less than 3 mm. Therefore, all adjustments were in the cephalad direction (due to caudal displacement). Table 2 shows the range of shift magnitudes.
Table 2 Occurrence of displacement magnitude. Magnitude (mm)
Number of implants
3 to 0.1 0 to 4.9 5 to 9.9 10 to 14.9 >15
3 13 18 9 5
The frequency of occurrence for each group of displacement magnitude is shown here. The number of implants refers to the frequency of occurrence.
The median time between acquisition of the planning CT scout and the pre-treatment X-ray was 254 min (range 81–367 min). Fig. 3 displays the elapsed time from the planning CT scout scan to pre-treatment X-ray and the magnitude of catheter displacement. The wide range in time from planning scan to treatment (81–367 min) was due to departmental logistics as the treatment room is also used to treat gynaecological malignancies with HDR brachytherapy. The catheter hub to template distance was measured for the 4 reference catheters after implantation with the legs in the lithotomy position, again prior to the CT planning scan with the legs down, and immediately prior to the pre-treatment verification X-ray with the legs down. The median displacement between each incidence of measurement was 0 mm (range 6–4 mm). A negative displacement indicates a cephalad shift of the reference catheter. Only one patient showed significant (>5 mm) external displacement in one reference catheter. The median variation in displacement between the reference catheters for the 3 separate measurement times was 0 mm (range 3–3 mm). With the exception of 2 patients, the variation in displacement between the 4 catheters was always less than 1 mm. We felt that applying the average displacement from the 4 reference catheters to adjust and compensate for the remaining catheters was valid. The distance of each fiducial marker respective to the other 2 markers was measured from the CT scout scan and the
M. Whitaker et al. / Radiotherapy and Oncology 101 (2011) 490–494
pre-treatment X-ray for each patient at each fraction. The median displacement between the CT scout scan and pre-treatment X-ray was 1 mm (range 0–8 mm). Only 3 patients showed significant (>5 mm) displacement of any markers during the elapsed time. We used the ANOVA technique to analyse the following factors and their first order interactions on increased displacement: (i) the number of catheters implanted, (ii) the time elapsed between planning CT scan and treatment, and, (iii) the time elapsed together with the number of catheters were considered. The results indicate that the number of catheters implanted (p = 0.01) and the interaction between number of catheters and elapsed time (p = 0.06) were statistically significant at a 10% level, while elapsed time (p = 0.12) itself is not. The correlation between patient fractions was not analysed as each patient only received 2 fractions and any correlation, or lack thereof, could not be statistically significant.
Discussion Catheter displacement between fractions has been well documented [1,4–6], however, shift occurring between the time of treatment planning and delivery has not. Our aim was to assess catheter displacement between treatment planning and delivery as in our HDR brachytherapy programme we deliver one fraction per implant. The principal finding was that catheters moved a median of 7.5 mm relative to the prostate in a median time of 254 min. Thirty-two of 48 implants (67%) underwent caudal displacement of 5 mm or more. Furthermore, there was caudal shift of 16 mm or more in 5 cases (10%). Catheter displacement is thought to be due to both acute oedema between the prostate and the perineal skin [1] as well as altered patient positioning due to removal of the ultrasound probe and then the lowering of the legs from a dorsal lithotomy position. The latter certainly results in changes to patient anatomy and implant geometry [7], and this may be accentuated by the transfer of the patient from the operating theatre bed to the patient bed and then onto the CT couch. However, in our facility patients undergo CT scan and treatment with their legs in the same position so any change in anatomy caused by lowering the patient’s legs occurs before treatment planning. Generally we found negligible changes in our repeat template to catheter hub measurements despite large displacements of catheters relative to fiducial markers, supporting Hoskin’s hypothesis of acute oedema as the major cause of implant displacement. We observed only one significant (>5 mm) caudal shift between the time of implantation and the planning CT scan, as measured from the cather hub to template. Our practice of suturing the template to the patient and activating the template’s built-in locking mechanism ensures minimal template and catheter slippage. With respect to delivering the treatment to the target as planned, it is the catheter displacement between the planning CT acquisition to the time of treatment delivery that is critical. Our results demonstrate that even the relatively short period of time between acquisition of the CT scout scan and pre-treatment X-ray is sufficient to observe the occurrence of oedema, and hence caudal catheter shift, in many patients. In this report, the occurrence and magnitude of the shift varied from implant to implant for each patient, and also from patient to patient. Other studies have reported similar magnitudes of displacement for inter-fraction motion [1,4,5]. In this study we compensated for caudal catheter displacement by digitally editing the indexer positions at the treatment console to accommodate for the displacements prior to treatment delivery. As discussed in the results section, the actioned shift was always in the cephalad direction. For this report, actioned shifts were those of 5 mm or more in magnitude based on Hoskin’s work.
493
One limitation of this study is that no reference catheters were defined in the posterior section. Various artefacts and anatomy render the mid-lateral catheters difficult to observe for some patients, hence the posterior positions were not considered. Another potential limitation of the displacement measurement is the reliance on AP X-ray images only, as the image quality from orthogonal X-rays was too poor to be used for clinical measurements. Furthermore, the CT scan scout image was used for measurement and comparison rather than a digitally reconstructed radiograph which would take beam divergence into account. The assumption that the catheters are displaced in the longitudinal axis only is not unreasonable, given that a radial shift of up to 20° will result in an error of 1 mm. Tiong et al. (2010) assume catheters displacement in the longitudinal axis only, performing a simple visual quality assurance to ensure minimal rotation between the CT scout image and AP X-ray, which we have adopted. We also assumed that individual catheters do not move relative to each other but displace as a unit relative to the prostate, based on the work by Tiong et al. (2010). A study by Kolkman-Deurloo et al. (2011) simulated the effect of displacement of catheters in selected template rows due to pelvic tilt. They found that the higher weighted dorsal displacements were more likely to affect the target coverage, as ventral rows tended to have lower weighting. However, due to the limitations of our current imaging system, we are unable to obtain reliable lateral X-rays with which to ascertain tilt, and, therefore, action any shifts to the entire implant without prejudice. An alternative strategy to measuring catheter displacement is the use of commercially available technologies to minimise catheter movement, such as implanting a deployable stabilising mechanism within certain catheters to anchor the neighbouring catheters within the prostate, rendering the catheter/template entity less likely to be displaced by oedema or patient movement [8]. Catheters such as the CookÒ Localization Needle (Cook Medical, Bloomington, USA), feature a hook mechanism which, when deployed, grips the prostate tissue. Using several of these catheters may assist in stabilising the implant within the prostate. Online ultrasound planning at the time of implantation has been investigated as a means of obviating the need for a CT scan. If this can be performed expediently, and with adequate image quality to clearly delineate the prostatic contours, this may circumvent, or at least, reduce the problem of catheter displacement. The use of ultrasound-based planning has been investigated by a Dutch group for the first fraction of HDR monotherapy [9]. However, this group showed that the ultrasound acquired treatment plan could not be reliably applied to subsequent fractions as the changes in anatomy, both from removal of the transrectal ultrasound probe and posture resulting from lowering the legs from the dorsal lithotomy position, had an adverse dosimetric effect, particularly on the urethra, resulting in 32% of urethral volume receiving P120% of the prescribed dose [9]. Tiong et al. (2010) reported significant adverse effects on the tumour control probability with displacements greater than 3 mm [2]. Caudal catheter migration of 3 mm or greater, relative to the prostate, may result in underdosage to the gland and overdosage to critical structures according to the D90 values or the COIN index [1,10], and also as evidenced in the study by Kolkman-Deurloo et al. (2011). Dose reductions of up to 27% in the D90 have been reported in cases without catheter movement correction, whereas less than 6% reduction of the D90 has been reported with movement correction between fractions [4]. It is not entirely clear whether catheter displacement in the Tiong study was measured prior to the first fraction of treatment or in fact represented inter-fraction displacement. This report highlights that even small displacements (be they between planning and treatment or between fractions) have significant effects on both tumour control probability and dose to organs at risk. Since the report of Tiong
494
Prostate brachytherapy
et al. (2010), we currently action any displacement of 3 mm or more. Sullivan et al. (2009) have demonstrated that urethral strictures are the predominant late urinary toxicity of prostate HDR brachytherapy, with a 12% actuarial risk of grade 2 or more stricture development at 6 years for HDR boost patients, and 15% at 3 years for HDR monotherapy patients. The median time to diagnosis of stricture was 22 months. 92% of all strictures occurred in the bulbo-membranous urethra, which lies distal to the prostatic apex [3]. With caudal displacement of magnitudes ranging up to 23 mm, it is highly likely that the high dose to the gland may be erroneously delivered to the bulbo-membranous urethra. Bulbo-membranous urethral stricture formation may, therefore, be a clinical manifestation of (uncorrected) caudal catheter displacement. Given the temporal delay to stricture formation, we plan to review our patient cohort to see if stricture rates are reduced by compensating for catheter displacement. Late rectal complications are unusual in the HDR boost setting, however, with long-term follow-up and a large enough patient cohort one could speculate that catheter displacement might also contribute to rectal toxicity as the rectum has a slight anterior deviation at the prostatic apex, bringing it closer to the target volume. Conclusions Displacement of 5 mm or more occurred in 67% of implants between CT planning and treatment delivery. We recommend that departments performing HDR prostate brachytherapy verify internal catheter positions immediately prior to any treatment delivery as well as between fractions in regimes where this is indicated. Although this may seem a logical step in any well-designed HDR brachytherapy workflow, many time poor departments do not undertake this basic quality assurance to the detriment of the patient. Conflict of interest statement None.
Acknowledgements We would like to acknowledge the brachytherapy team at RPAH department of Radiation Oncology for their assistance with data collection, and Mr. Jonathan Whitaker for his assistance with the statistical analysis. References [1] Hoskin P, Bownes P, Ostler P, Walker K, Bryant L. High dose rate afterloading brachytherapy for prostate cancer: catheter and gland movement between fractions. Radiother Oncol 2003;68:285–8. [2] Tiong A, Bydder S, Ebert M, Caswell N, Waterhouse D, Spry N, et al. A small tolerance for catheter displacement in high-dose rate prostate brachytherapy is necessary and feasible. Int J Radiat Oncol Biol Phys 2010;76: 1066–72. [3] Sullivan L, Williams S, Tai K, Foroudi F, Cleeve L, Duchesne G. Urethral structure following high dose rate brachytherapy for prostate cancer. Radiother Oncol 2009;91:232–6. [4] Simnor T, Li S, Lowe G, Ostler P, Bryant L, Chapman C, et al. Justification for inter-fraction correction of catheter movement in fractionated high dose-rate brachytherapy treatment of prostate cancer. Radiother Oncol 2009;93: 253–8. [5] Damore S, Syed N, Puthawala A, Sharma A. Needle displacement during HDR brachytherapy in the treatment of prostate cancer. Int J Radiat Oncol Biol Phys 2000;46:1205–11. [6] Kim Y, Hsu I-C, Pouliot J. Measurement of craniocaudal catheter displacement between fractions in computed tomography-based high dose rate brachytherapy of prostate cancer. J Appl Clin Med Phys 2007;8:1–13. [7] Kolkman-Deurloo I, de Langen M. Ultrasound-based versus CT-based HDR brachytherapy of the prostate: the influence of anatomic differences on dosimetry. Radiother Oncol 2004;73:S238–9. [8] Kolkman-Deurloo I, Roos M, Aluwini S. HDR monotherapy for prostate cancer: a simulation study to determine the effect of catheter displacement on target coverage and normal tissue irradiation. Radiother Oncol 2011;98: 192–7. [9] Seppenwoolde Y, Kolkman-Deurloo I-K, Sipkema D, dL M, Praag J, Jansen P, et al. HDR prostate monotherapy – Dosimetric effects of implant deformation due to posture change between TRUS- and CT-imaging. Radiother Oncol 2008;86:114–9. [10] Tiong A, Bydder S, Ebert M, Caswell N, Waterhouse D, Spry N, et al. A small tolerance for catheter displacement in high-dose rate prostate brachytherapy is necessary and feasible. Int J Radiat Oncol Biol Phys 2010;86:1066–72.
Contents lists available at SciVerse ScienceDirect
Radiotherapy and Oncology journal homepage: www.thegreenjournal.com
Prostate cancer brachytherapy
Prostate HDR brachytherapy catheter displacement between planning and treatment delivery May Whitaker a,⇑, George Hruby a,b, Aimee Lovett a, Nitya Patanjali a a
Department of Radiation Oncology, Sydney Cancer Centre; and b Division of Medicine, University of Sydney, NSW, Australia
a r t i c l e
i n f o
Article history: Received 27 October 2010 Received in revised form 3 August 2011 Accepted 5 August 2011 Available online 31 August 2011 Keywords: HDR prostate brachytherapy Catheter displacement Movement
a b s t r a c t Background and purpose: HDR brachytherapy is used as a conformal boost for treating prostate cancer. Given the large doses delivered, it is critical that the volume treated matches that planned. Our outpatient protocol comprises two 9 Gy fractions, two weeks apart. We prospectively assessed catheter displacement between CT planning and treatment delivery. Materials and methods: Three fiducial markers and the catheters were implanted under transrectal ultrasound guidance. Metal marker wires were inserted into 4 reference catheters before CT; marker positions relative to each other and to the marker wires were measured from the CT scout. Measurements were repeated immediately prior to treatment delivery using pelvic X-ray with marker wires in the same reference catheters. Measurements from CT scout and film were compared. For displacements of 5 mm or more, indexer positions were adjusted prior to treatment delivery. Results: Results are based on 48 implants, in 25 patients. Median time from planning CT to treatment delivery was 254 min (range 81–367 min). Median catheter displacement was 7.5 mm (range 2.9– 23.9 mm), 67% of implants had displacement of 5 mm or greater. Displacements were predominantly caudal. Conclusions: Catheter displacement can occur in the 1–3 h between the planning CT scan and treatment. It is recommended that departments performing HDR prostate brachytherapy verify catheter positions immediately prior to treatment delivery. Crown Copyright Ó 2011 Published by Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology 101 (2011) 490–494
In high dose rate (HDR) brachytherapy the radiation dose decreases rapidly outside the treatment volume, making it an ideal technique for treating targets lying close to organs at risk such as the prostate gland. With imaging tools such as ultrasound and CT, together with a 3D treatment planning system and anatomical contouring, dose volume histograms are relatively quick and simple to generate. However, it is critical that the treated volume matches the planned volume due to the large doses delivered per fraction, otherwise there is a risk of over-treating critical organs such as the urethra or rectum, and under-treating the target. This has been demonstrated in work at Mt Vernon Cancer Centre in the UK [1] and Sir Charles Gairdner Hospital in Australia [2]. Clinically, caudal displacement of catheters may be responsible for the relatively high incidence of urethral strictures seen after HDR brachytherapy [3]. In this study, we prospectively assessed catheter movement between time of CT imaging and delivery of the first fraction. ⇑ Corresponding author. Address: Royal Prince Alfred Hospital, Department of Radiation Oncology, Building 27 Missenden Rd., Camperdown, NSW 2050, Australia. E-mail address: [email protected] (M. Whitaker).
Materials and methods Patient selection criteria Patient selection for HDR prostate brachytherapy boost included suitability for anaesthesia, life expectancy of at least 10 years, adequate baseline urinary function, and prostate volume less than 60 cc. Tumour characteristics included serum PSA of 10 lg/l or more; and/or Gleason Score of 7 or more; and no evidence of metastases see Table 1. Tumour treatment At the Sydney Cancer Centre (SCC), HDR brachytherapy is used in the management of prostate cancer as a conformal boost in conjunction with external beam radiation treatment (EBRT). The boost is delivered as two fractions of 9 Gy, two weeks apart, with EBRT of 46 Gy in 23 fractions starting approximately 4–5 days after the first brachytherapy treatment. A one day break in EBRT occurs on the day of the second brachytherapy implant and treatment. This regimen has been adopted for patient comfort, to avoid hospitalization, and to eliminate any inter-fraction implanted catheter movement.
0167-8140/$ - see front matter Crown Copyright Ó 2011 Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.radonc.2011.08.004
M. Whitaker et al. / Radiotherapy and Oncology 101 (2011) 490–494 Table 1 Tumour characteristics of the 25 patients involved in this study. PSA
<10 10–20 >20
8 15 2
Tumour stage
T1 T2 T3
7 17 1
Gleason score
6 3+4 4+3 8
1 10 13 1
No: 4
Yes: 21
Androgen deprivation
The implant procedures took place in our departmental treatment bunker which is operating-theatre equipped. All patients underwent either general or spinal anaesthesia and were placed in the lithotomy position. After insertion of an in-dwelling urinary catheter, 3 fiducial markers (CIVCO ACCULOCÒ, Orange City, USA) were placed transperineally into the prostate (2 at the base, 1 at mid or apex) under ultrasound guidance using a prostate stepper template (Nucletron, Veenendaal, Netherlands), at the first implant procedure only. Subsequently, the Pro-guide catheters (Nucletron, Veenendaal, Netherlands) were placed using the same template which was firmly sutured to the perineum at the end of the procedure (Fig. 1). The catheters were implanted to well beyond the base of the prostate (approximately 2 cm into the bladder) to allow for compensation of any displacement measured before treatment delivery. Routine cystoscopy was not performed, however, catheter incursion into the bladder was documented with sagittal ultrasound imaging. Once the catheters were labelled, and the distance from template to catheter hubs measured, the patient’s legs were removed from the stirrups and the patient was transferred (down a short corridor) to our CT scanner. The patient was positioned on the CT couch with his legs down, and the distance from template to catheter hubs was again measured and confirmed prior to scanning. The patient was then placed back on the hospital bed to await final transfer to the treatment room for dose delivery after treatment planning. Treatment planning was performed using the PLATO planning system (Nucletron, Veenendaal, Netherlands). The Radiation Oncologists (GH and NP) contoured the target and organs at risk, and the PLATO Inverse Planning Simulated Annealing algorithm was used to optimise the dwell positions and times.
Fig. 1. Catheters implanted in a patient using a prostate stepper template. The template is sutured to the patient’s skin and the template to catheter hub distance is measured.
491
Prior to treatment delivery, the distance from template to catheter hubs was measured a third and final time. An anterior– posterior (AP) digital X-ray image was taken with the patient’s legs down, and treatment was delivered in this position on the hospital bed. Each fraction was delivered several hours after implantation. The template was removed immediately following treatment, with the IDC being removed 2–3 h later following bladder irrigation. Equipment Equipment for all cases included the following – Microselectron Iridium-192 afterloader, PLATO planning software, Pro-guide catheters, Microtouch stepper-stabilizer unit (all Nucletron, Veenendaal, Netherlands) and B&K Ultrasound unit (B&K, Herlev, Denmark). The template to proximal catheter hub distance for 4 reference catheters was measured after each insertion with the legs still in the lithotomy position, to establish a baseline reading. This measurement was repeated with the legs down at the CT planning scan and again immediately prior to treatment. Planning was undertaken using a CT scan (Toshiba, Tustin, USA) with 3 mm slice thickness and the PLATO planning system. To compare catheter positions at the time of CT scanning and time of treatment delivery, metal marker wires were inserted into the same 4 reference catheters at CT and during the digital X-ray acquisition; seed positions relative to each other (measuring from the centre of each seed) and vertical measurement from the tip of each of the 4 marker wires were recorded from the CT scout image and the pre-treatment X-ray. Reference catheters were defined as the two most anterior/medial catheters and the two most lateral catheters. Seed to seed measurements as well as catheter to seed measurements were used to assess any implant displacement relative to the prostate. See Fig. 2. Measurement of displacement Internal seed and reference marker wire position measurements were repeated immediately prior to treatment delivery using an AP pelvic X-ray with the marker wires again inserted in the same 4 reference catheters used at the time of planning scan. Measurements from the CT scout and film were compared and the average value over the 4 reference catheters from each image was used as the displacement magnitude (Fig. 3).
Fig. 2. Measurement process for seed to seed and seed to catheter measurement. The distance from the tip of the catheter to the tip of all 3 fiducial markers is measured for the 4 catheters. These values are then averaged to produce a single value for the catheter to seed distance for the planning CT scout. The process is then repeated using the pre-treatment X-ray. The difference between the two values is recorded as the measured displacement.
492
Prostate brachytherapy
25.0
Measured shift distance (mm)
20.0
15.0
10.0
5.0
0.0
7 33
9
1 33
0 31
31
5
3 30
30
0
7 28
28
4
8 26
27
7 25
1
4 23
25
5
9 19
4 19
19
4
0 19
17
9
5 16
15
7
1 15
12
81
-5.0
Time elapsed between acquisition of CT scan and acquisition of digital x-ray (min) Fig. 3. The time elapsed between acquisition of the CT scout scan and the pre-treatment X-ray does not appear to be a consideration in the magnitude of catheter displacement.
It was assumed that catheter displacements occurred only in the longitudinal axis [2]. A radial shift of up to 20° will result in an error of approximately 1 mm in the displacement measurement. This was considered insignificant with respect to the method of displacement calculation. Orthogonal X-rays were initially attempted, however, accurate measurements could not be made due to the poor image quality; therefore, AP X-rays alone were used for displacement measurements. Shift protocol Based on the action threshold used at Mt. Vernon [4], displacements of 5 mm or more measured between planning CT and pretreatment X-ray were actioned. To compensate for any measured displacement greater than 5 mm, the source was digitally advanced further into the patient by altering the indexer length at the treatment console. Implanting the catheters past the prostate base into the bladder allowed for extra dwell positions beyond the prostate in the event of caudal shift. Physical re-insertion was not performed. Results The 25 patients in this study had a median age of 67 yr (range 47–78). Most men (n = 22) had intermediate risk prostate cancer. Three men had high risk disease, of whom two had 1 adverse feature each and one patient had 2 adverse features (T3 and PSA > 20). The median PSA was 12.4 (range 4–32). The tumour characteristics are detailed in Table 1. Our results are based on a total of 48 implants performed on 25 patients, with a median of 18 catheters (range 15–26) per implant. The median measured displacement was 7.5 mm (range 2.9– 23.9 mm), with shifts of greater than or equal to 5 mm occurring in 32 implants (67%). Displacement in the cephalad direction (as represented by a negative value) was uncommon (3 of 48 cases only) and was always less than 3 mm. Therefore, all adjustments were in the cephalad direction (due to caudal displacement). Table 2 shows the range of shift magnitudes.
Table 2 Occurrence of displacement magnitude. Magnitude (mm)
Number of implants
3 to 0.1 0 to 4.9 5 to 9.9 10 to 14.9 >15
3 13 18 9 5
The frequency of occurrence for each group of displacement magnitude is shown here. The number of implants refers to the frequency of occurrence.
The median time between acquisition of the planning CT scout and the pre-treatment X-ray was 254 min (range 81–367 min). Fig. 3 displays the elapsed time from the planning CT scout scan to pre-treatment X-ray and the magnitude of catheter displacement. The wide range in time from planning scan to treatment (81–367 min) was due to departmental logistics as the treatment room is also used to treat gynaecological malignancies with HDR brachytherapy. The catheter hub to template distance was measured for the 4 reference catheters after implantation with the legs in the lithotomy position, again prior to the CT planning scan with the legs down, and immediately prior to the pre-treatment verification X-ray with the legs down. The median displacement between each incidence of measurement was 0 mm (range 6–4 mm). A negative displacement indicates a cephalad shift of the reference catheter. Only one patient showed significant (>5 mm) external displacement in one reference catheter. The median variation in displacement between the reference catheters for the 3 separate measurement times was 0 mm (range 3–3 mm). With the exception of 2 patients, the variation in displacement between the 4 catheters was always less than 1 mm. We felt that applying the average displacement from the 4 reference catheters to adjust and compensate for the remaining catheters was valid. The distance of each fiducial marker respective to the other 2 markers was measured from the CT scout scan and the
M. Whitaker et al. / Radiotherapy and Oncology 101 (2011) 490–494
pre-treatment X-ray for each patient at each fraction. The median displacement between the CT scout scan and pre-treatment X-ray was 1 mm (range 0–8 mm). Only 3 patients showed significant (>5 mm) displacement of any markers during the elapsed time. We used the ANOVA technique to analyse the following factors and their first order interactions on increased displacement: (i) the number of catheters implanted, (ii) the time elapsed between planning CT scan and treatment, and, (iii) the time elapsed together with the number of catheters were considered. The results indicate that the number of catheters implanted (p = 0.01) and the interaction between number of catheters and elapsed time (p = 0.06) were statistically significant at a 10% level, while elapsed time (p = 0.12) itself is not. The correlation between patient fractions was not analysed as each patient only received 2 fractions and any correlation, or lack thereof, could not be statistically significant.
Discussion Catheter displacement between fractions has been well documented [1,4–6], however, shift occurring between the time of treatment planning and delivery has not. Our aim was to assess catheter displacement between treatment planning and delivery as in our HDR brachytherapy programme we deliver one fraction per implant. The principal finding was that catheters moved a median of 7.5 mm relative to the prostate in a median time of 254 min. Thirty-two of 48 implants (67%) underwent caudal displacement of 5 mm or more. Furthermore, there was caudal shift of 16 mm or more in 5 cases (10%). Catheter displacement is thought to be due to both acute oedema between the prostate and the perineal skin [1] as well as altered patient positioning due to removal of the ultrasound probe and then the lowering of the legs from a dorsal lithotomy position. The latter certainly results in changes to patient anatomy and implant geometry [7], and this may be accentuated by the transfer of the patient from the operating theatre bed to the patient bed and then onto the CT couch. However, in our facility patients undergo CT scan and treatment with their legs in the same position so any change in anatomy caused by lowering the patient’s legs occurs before treatment planning. Generally we found negligible changes in our repeat template to catheter hub measurements despite large displacements of catheters relative to fiducial markers, supporting Hoskin’s hypothesis of acute oedema as the major cause of implant displacement. We observed only one significant (>5 mm) caudal shift between the time of implantation and the planning CT scan, as measured from the cather hub to template. Our practice of suturing the template to the patient and activating the template’s built-in locking mechanism ensures minimal template and catheter slippage. With respect to delivering the treatment to the target as planned, it is the catheter displacement between the planning CT acquisition to the time of treatment delivery that is critical. Our results demonstrate that even the relatively short period of time between acquisition of the CT scout scan and pre-treatment X-ray is sufficient to observe the occurrence of oedema, and hence caudal catheter shift, in many patients. In this report, the occurrence and magnitude of the shift varied from implant to implant for each patient, and also from patient to patient. Other studies have reported similar magnitudes of displacement for inter-fraction motion [1,4,5]. In this study we compensated for caudal catheter displacement by digitally editing the indexer positions at the treatment console to accommodate for the displacements prior to treatment delivery. As discussed in the results section, the actioned shift was always in the cephalad direction. For this report, actioned shifts were those of 5 mm or more in magnitude based on Hoskin’s work.
493
One limitation of this study is that no reference catheters were defined in the posterior section. Various artefacts and anatomy render the mid-lateral catheters difficult to observe for some patients, hence the posterior positions were not considered. Another potential limitation of the displacement measurement is the reliance on AP X-ray images only, as the image quality from orthogonal X-rays was too poor to be used for clinical measurements. Furthermore, the CT scan scout image was used for measurement and comparison rather than a digitally reconstructed radiograph which would take beam divergence into account. The assumption that the catheters are displaced in the longitudinal axis only is not unreasonable, given that a radial shift of up to 20° will result in an error of 1 mm. Tiong et al. (2010) assume catheters displacement in the longitudinal axis only, performing a simple visual quality assurance to ensure minimal rotation between the CT scout image and AP X-ray, which we have adopted. We also assumed that individual catheters do not move relative to each other but displace as a unit relative to the prostate, based on the work by Tiong et al. (2010). A study by Kolkman-Deurloo et al. (2011) simulated the effect of displacement of catheters in selected template rows due to pelvic tilt. They found that the higher weighted dorsal displacements were more likely to affect the target coverage, as ventral rows tended to have lower weighting. However, due to the limitations of our current imaging system, we are unable to obtain reliable lateral X-rays with which to ascertain tilt, and, therefore, action any shifts to the entire implant without prejudice. An alternative strategy to measuring catheter displacement is the use of commercially available technologies to minimise catheter movement, such as implanting a deployable stabilising mechanism within certain catheters to anchor the neighbouring catheters within the prostate, rendering the catheter/template entity less likely to be displaced by oedema or patient movement [8]. Catheters such as the CookÒ Localization Needle (Cook Medical, Bloomington, USA), feature a hook mechanism which, when deployed, grips the prostate tissue. Using several of these catheters may assist in stabilising the implant within the prostate. Online ultrasound planning at the time of implantation has been investigated as a means of obviating the need for a CT scan. If this can be performed expediently, and with adequate image quality to clearly delineate the prostatic contours, this may circumvent, or at least, reduce the problem of catheter displacement. The use of ultrasound-based planning has been investigated by a Dutch group for the first fraction of HDR monotherapy [9]. However, this group showed that the ultrasound acquired treatment plan could not be reliably applied to subsequent fractions as the changes in anatomy, both from removal of the transrectal ultrasound probe and posture resulting from lowering the legs from the dorsal lithotomy position, had an adverse dosimetric effect, particularly on the urethra, resulting in 32% of urethral volume receiving P120% of the prescribed dose [9]. Tiong et al. (2010) reported significant adverse effects on the tumour control probability with displacements greater than 3 mm [2]. Caudal catheter migration of 3 mm or greater, relative to the prostate, may result in underdosage to the gland and overdosage to critical structures according to the D90 values or the COIN index [1,10], and also as evidenced in the study by Kolkman-Deurloo et al. (2011). Dose reductions of up to 27% in the D90 have been reported in cases without catheter movement correction, whereas less than 6% reduction of the D90 has been reported with movement correction between fractions [4]. It is not entirely clear whether catheter displacement in the Tiong study was measured prior to the first fraction of treatment or in fact represented inter-fraction displacement. This report highlights that even small displacements (be they between planning and treatment or between fractions) have significant effects on both tumour control probability and dose to organs at risk. Since the report of Tiong
494
Prostate brachytherapy
et al. (2010), we currently action any displacement of 3 mm or more. Sullivan et al. (2009) have demonstrated that urethral strictures are the predominant late urinary toxicity of prostate HDR brachytherapy, with a 12% actuarial risk of grade 2 or more stricture development at 6 years for HDR boost patients, and 15% at 3 years for HDR monotherapy patients. The median time to diagnosis of stricture was 22 months. 92% of all strictures occurred in the bulbo-membranous urethra, which lies distal to the prostatic apex [3]. With caudal displacement of magnitudes ranging up to 23 mm, it is highly likely that the high dose to the gland may be erroneously delivered to the bulbo-membranous urethra. Bulbo-membranous urethral stricture formation may, therefore, be a clinical manifestation of (uncorrected) caudal catheter displacement. Given the temporal delay to stricture formation, we plan to review our patient cohort to see if stricture rates are reduced by compensating for catheter displacement. Late rectal complications are unusual in the HDR boost setting, however, with long-term follow-up and a large enough patient cohort one could speculate that catheter displacement might also contribute to rectal toxicity as the rectum has a slight anterior deviation at the prostatic apex, bringing it closer to the target volume. Conclusions Displacement of 5 mm or more occurred in 67% of implants between CT planning and treatment delivery. We recommend that departments performing HDR prostate brachytherapy verify internal catheter positions immediately prior to any treatment delivery as well as between fractions in regimes where this is indicated. Although this may seem a logical step in any well-designed HDR brachytherapy workflow, many time poor departments do not undertake this basic quality assurance to the detriment of the patient. Conflict of interest statement None.
Acknowledgements We would like to acknowledge the brachytherapy team at RPAH department of Radiation Oncology for their assistance with data collection, and Mr. Jonathan Whitaker for his assistance with the statistical analysis. References [1] Hoskin P, Bownes P, Ostler P, Walker K, Bryant L. High dose rate afterloading brachytherapy for prostate cancer: catheter and gland movement between fractions. Radiother Oncol 2003;68:285–8. [2] Tiong A, Bydder S, Ebert M, Caswell N, Waterhouse D, Spry N, et al. A small tolerance for catheter displacement in high-dose rate prostate brachytherapy is necessary and feasible. Int J Radiat Oncol Biol Phys 2010;76: 1066–72. [3] Sullivan L, Williams S, Tai K, Foroudi F, Cleeve L, Duchesne G. Urethral structure following high dose rate brachytherapy for prostate cancer. Radiother Oncol 2009;91:232–6. [4] Simnor T, Li S, Lowe G, Ostler P, Bryant L, Chapman C, et al. Justification for inter-fraction correction of catheter movement in fractionated high dose-rate brachytherapy treatment of prostate cancer. Radiother Oncol 2009;93: 253–8. [5] Damore S, Syed N, Puthawala A, Sharma A. Needle displacement during HDR brachytherapy in the treatment of prostate cancer. Int J Radiat Oncol Biol Phys 2000;46:1205–11. [6] Kim Y, Hsu I-C, Pouliot J. Measurement of craniocaudal catheter displacement between fractions in computed tomography-based high dose rate brachytherapy of prostate cancer. J Appl Clin Med Phys 2007;8:1–13. [7] Kolkman-Deurloo I, de Langen M. Ultrasound-based versus CT-based HDR brachytherapy of the prostate: the influence of anatomic differences on dosimetry. Radiother Oncol 2004;73:S238–9. [8] Kolkman-Deurloo I, Roos M, Aluwini S. HDR monotherapy for prostate cancer: a simulation study to determine the effect of catheter displacement on target coverage and normal tissue irradiation. Radiother Oncol 2011;98: 192–7. [9] Seppenwoolde Y, Kolkman-Deurloo I-K, Sipkema D, dL M, Praag J, Jansen P, et al. HDR prostate monotherapy – Dosimetric effects of implant deformation due to posture change between TRUS- and CT-imaging. Radiother Oncol 2008;86:114–9. [10] Tiong A, Bydder S, Ebert M, Caswell N, Waterhouse D, Spry N, et al. A small tolerance for catheter displacement in high-dose rate prostate brachytherapy is necessary and feasible. Int J Radiat Oncol Biol Phys 2010;86:1066–72.