INTRODUCTION
Recent studies have shown increased prostate cancer cure rates with dose escalation of radiation dose.Reference Pollack, Zagars, Starkschall, Antolak, Lee, Huang, von Eschenbach, Kuban and Rosen1,Reference Hanks, Schultheiss, Hanlon, Hunt, Lee, Epstein and Coia2 However, in order to maximize dose to the prostate, it is necessary to minimize dose to rectum and bladder. Three dimensional conformal radiotherapy (3D-CRT) and intensity modulated radiotherapy (IMRT) has shown better results in terms of delivering higher radiation doses to the prostate and acceptable acute and late genitourinary (GU) and gastrointestinal (GI) toxicity.Reference Nilsson, Norlén and Widmark3,Reference Zelefsky, Fuks and Happersett4 However, some of the issues, for example uncertainties regarding target localization, organ motion and setup errors, may offset the advantages of 3D-CRT and IMRT.Reference Langen and Jones5
High-dose-rate brachytherapy (HDR) is an alternative method to deliver higher dose to the target as with needles in place only insignificant movements occur during treatment. Furthermore, a steep dose gradient between target and adjacent organs at risk spares the surrounding organs from being exposed to an excessive radiation dose despite the delivery of a higher dose to the prostate.Reference Kälkner, Wahlgren and Ryberg6
However, similar to setup errors in external beam radiotherapy (EBRT), various factors may cause the dose uncertainties in HDR brachytherapy, including needle induced trauma and subsequent change in target volume,Reference Kim, Hsu, Lessard, Vujic and Pouliot7 edema resolution between fractions,Reference Mullokandov and Gejerman8 visualization of urethra Reference Bucci, Spadinger, Hilts, Sidhu, Smith, Keyes and Morris9 and computed tomography (CT) slice thickness.Reference Kim, Hsu, Lessard, Pouliot and Vujic10 The desired position or depth of needles is the key point for successful dose delivery and is highly dependent on operator’s experience. Higher number of attempts per catheter may result in increased trauma along the needle track and may cause increased GU toxicity. After the implantation, individual needles do not move relative to one another or to perineal templates used, but there is a tendency for them to displace inferiorly relative to the prostate as a single unit,Reference Damore, Syed, Puthawala and Sharma11 this can lead to under dosage of prostate. Various methods have been proposed to compensate this movement.Reference Pellizzon, Salvajoli, Novaes, Maia, Ferigno and Fogaroli12
Our study aimed to evaluate the impact of catheters position, their number and repeated catheter manipulations on the optimal dose to the prostate and urethra, tumour control probability (TCP) and on the severity of acute GU toxicity; subsequently these results were used to find the solutions for adequate position, number of the catheters and various dosimetric parameters.
MATERIALS AND METHODS
Patients and treatment techniques
Fifty consecutive patients, after written consent for the treatment, received 192-iridium (Ir192) HDR brachytherapy boost to prostate along with 3D-CRT between July 2008 to August 2009. Study protocol was approved by institutional ethical committee before patients’ accrual. The characteristics of the patients are shown in Table 1. Briefly, conformal radiotherapy was administrated with high-energy photons of 15MV X-rays at total dose of 46 Gy. Total dose was given in five times weekly with fraction doses of 2 Gy. All patients were treated by conformal radiotherapy using multileaf collimators. HDR brachytherapy was given in two sessions at 20 and 40 Gy of EBRT. In a dedicated room, keeping patient in lithotomy position under spinal anaesthesia, urologists inserted needles transperineally into the prostate guided by transrectal ultrasonography cranial images after identifying contrast filled urethra. After all the procedure was completed, real-time images were obtained and all organs (prostate, urethra, bladder and rectum) were contoured with help of radiologist, and treatment planning was done on Flexiplan version 2.2. Dwell positions were activated at 2.5 mm along each needle. Dwell times were optimized using reverse planning optimization algorithm. The prescribed dose was 14 Gy at periphery of prostate and the mean prostate dose was 10 Gy; urethral dose constraints were kept low so that mean urethral dose does not exceed beyond 10 Gy. All data were collected during first session of brachytherapy.
Patient characteristics
PSA = prostate-specific antigen; EBRT = external beam radiotherapy.
All patients underwent androgen blockade prior to radiotherapy, which was continued with and after radiotherapy.
Number and attempts per catheter
During the procedure, the total catheters implanted were counted and also number of attempts per catheter to achieve desired position was recorded. Correlation analysis was done between number of catheters and different prostate volumes.
Mean catheters distance
The base of urinary bladder was taken as surrogate for the base of prostate (extreme caudal part of prostate). Thereafter, actual distances between the tips of all catheters and the base of bladder were measured separately on anteroposterior (AP) plain radiographs of pelvis with 12 inch C-arm unit (OEC 9800Plus; GE Medical Systems) (Figure 1) and on digitally reconstructed AP images from ultrasonography images of implanted catheters. The catheters were pushed in their optimal positions if distances were unacceptable. Catheter positions were again checked.
Mean tips of catheters to bladder distance.
Dose volume histogram (DVH) evaluation
The DVH derived from Flexiplan version 2.2 treatment planning system. The prostate volumes, urethra volumes, percentages of entire prostate and urethral volume receiving 10, 30, 50, 80, 100, 120 and 150% of prescribed dose (V10, V30, V50, V80, V100, V120 and V150) were recorded. Analysis was done between position of catheters and D90 (dose received by 90% of prostate volume).
Tumour control probability calculation (TCP):
The method for calculating TCP was based on linear quadratic (LQ) and TCP models by Wang et al.Reference Wang, Li, Yu and DiBiase13 Summarized equation is as follows:
Where e is exponential and K is cell number of tumour clonogens. The value of this parameter depends on the risk level of patient groups. For low, intermediate and high-risk prostate cancer, K values are 1.6 × 10Reference Kälkner, Wahlgren and Ryberg6, 3.0 × 10Reference Kälkner, Wahlgren and Ryberg6 and 1.1×10Reference Kim, Hsu, Lessard, Vujic and Pouliot7 cells. S is surviving fraction of cells irradiated to total dose. S was calculated by following equation according to LQ model.
Where D is the dose at which cells survive, and the dimensions of the parameters were α = Gy−1 and β = Gy−2 and the α/β ratio was considered 1.5 for prostate cancer. The interpretation of above equation is that cell killing results from the interaction of two elementary damaged species—probably DNA double strand breaks to produce dicentric chromosomal aberration which cause lethality.
The resultants TCP were plotted against catheters distances from the bladder (in our patients all calculation were done considering all patients with high risk).
Acute GU toxicity grading
Acute GU toxicities were graded on the toxicity criteria of Radiation Therapy Oncology Group and European Organization for Research and Treatment of Cancer (RTOG/EORTC CTC). Correlation and univariate analyses were done. Furthermore, the effect of number of attempts per catheters and their number on severity of acute GU toxicity was also evaluated.
Statistical analysis
All descriptive data, statistical tests correlation and chi-square test were analysed by using Statistical Package for the Social Sciences (SPSS) version 17.0.
RESULTS
Total number of catheters implanted and number of attempts per needle
The mean number of catheters implanted was 12.38 (range = 8–19), and number of attempts per catheter to optimize the position was 1.6 (range = 0–5). The mean prostate volume was 27.54 cc (10.8–69.0). No correlation was found between the prostate volume (p value 0.5) and number of catheters implanted (p value 0.3) (Figure 2).
Correlation between the prostate volume and number of catheters implanted.
Mean catheter distance and TCP analysis
Mean distance between the tips of catheters to base of bladder was 3.2 mm (1–8 mm) after adjustment. Before adjustment, mean distance was 6 mm. At 5 mm distance, D90 dose was 8.4 Gy prescribed dose, and for distances more than 5 mm, D90 dose dropped abruptly to 6 Gy at 8 mm (p < 0.01). Greater distances even showed lower D90 (Figure 3). This decrease in dose to the prostate resulted in lower relative TCP: 0.95, 0.95 and 0.66 for 3 mm, 5 mm and 6 mm distances of catheters to bladder (p < 0.01) (Figure 4).
D90 (dose received by 90% of prostate volume) of prostate decrement with increase in distance between tips of catheters and base of bladder.
Predicted tumour control probability decreases with increase in gap between catheters and bladder.
DVH data
The urethra is organ at risk (OAR) within the target, that is, prostate. Its position makes urethra to receive doses similar to prescribed doses. Mean urethral volume was 0.34 cc (0.1–0.8 cc). Mean urethral D90 was 7.5 Gy (5–11). No correlation was seen between very close positions of catheters to urethral doses, as dwell position and time were kept low or nil during treatment.
Acute GU toxicity and severity analysis
The main acute toxicities were dysuria, increased frequency and nocturia. The distribution of acute GU toxicity was as follows: Grade 0 = 58% (29 patients), Grade 1 = 20% (10 patients), Grade 2 = 14% (7 patients) and Grades 3 and 4 = 6% (1 patient). Severity of acute GU toxicity did not increase with number of catheters implanted, needling attempts. Similarly, no correlation was seen between the severity and the prostate volumes. The severity of acute GU toxicity was found to increase with urethral V10, V30, V50, V80 and V100 but did not show significance (p value = 0.9) until V120 and V130 were achieved (p value 0.001) (Figure 5).
Correlation between urethral doses with acute genitourinary toxicity.
DISCUSSION
Our study found no apparent correlation between number of catheters implanted and prostate volumes and acceptable dose distribution, but theoretically fewer needles may cause dose inhomogeneities. Study by Brunaud et al. showed a well-acceptable dose distribution for bigger prostates with fewer needles as compared to smaller prostates. Furthermore, this study found increased V150 with fewer needles. This may seem paradoxical at first, but it is related to the inverse square distance.Reference Charra-Brunaud, Hsu, Weinberg and Pouliot14
Mean distances between the tips of catheters to base of bladder strongly correlated with D90 and tumour-control probability. Distances more than 5 mm resulted in significant reduction of D90 and relative lower tumour-control probabilities.
Though our method of measuring the distances of tips of catheters from contrast filled bladder may carry some errors, due to ultrasonography-based online contouring and verification, negligible change in catheters orientation can occur as compared to CT-based reconstructed imaging and contouring. We did not use fiducial markers due to cost issues. As fiducial markers are inserted before catheter insertion, their position may change due to prostate stretching during needle implantation.Reference Kim, Hsu and Pouliot15 We did not measure the distances in lateral radiographs where catheter positions may change secondary to prostate edema, which is considered insignificant factor.
For calculating the TCP, we used biological parameters of Wang et al.; however, these vary widely. Our calculated TCP would not necessarily translate in lower disease-free survival, but minimizing the catheter distances from bladder by adjustment before treatment delivery can change the outcome. Recent study by Tiong et al., using similar parameters as we used in our study, found median relative TCP was 0.998, 0.964, 0.797 and 0.265 for caudal catheter displacements of 3, 6, 9 and 12 mm, respectively (p < 0.01 when all medians were compared).Reference Tiong, Bydder, Ebert, Caswell, Waterhouse, Spry, Camille and Joseph16
Regarding the factors influencing the severity of acute GU toxicity, the DVH data showed strong correlation between acute GU grading and urethral V120 and V130. The position of urethra itself makes it difficult to avoid delivering a higher dose. We made possible efforts to keep lower or zero dwell positions and dwell times for catheter which were adjacent to the urethra. Akimoto et al. also found that large volume of urethra receiving either prescribed dose or lesser may be related to severity of acute GU toxicity.Reference Akimoto, Ito and Saitoh17
We did not find any correlation between severity of acute GU toxicity and number of catheters implanted and number of attempts done per catheter to achieve desired position. One reason could be that we used inverse planning system to optimize the dose. Studies reported higher incidence of acute GU toxicity with fewer catheters used in geometric optimization algorithms. However, further study is required to determine the optimal number of catheters and attempts per catheter. Similarly, prostate volumes did not show any influence on severity of acute GU toxicity.
CONCLUSIONS
Following factors must be considered while planning a patient for HDR boost.
1. Mean catheter to base of bladder (or fiducial markers if used) distances shall be kept less than 5 mm. Increased distances reduce the dose distribution and relative TCP.
2. V120 to V130 of urethra shall be minimized to reduce the acute GU toxicity.
3. Upto five attempts per catheter insertion, we did not see any correlation with acute GU toxicity. Further study is required to see any impact of repeated attempts on GU toxicity.
4. Dwell positions and dwell times shall be minimized for catheters which are very close to urethra.
Conflicting interests
No potential conflict of interest, no grant received.
Acknowledgement
We thank to Dr. Adibul Hasan Rizvi, for his constant support and providing us state of art radiotherapy facilities used for cancer patients free of costs with dignity.
