- Email: [email protected]
81 oral BREAST TISSUE MODELING IN MONTE CARLO POST-IMPLANT EVALUATION OF BREAST LDR BRACHYTHERAPY
S 30
P ROFFERED PAPER
M ONDAY, M AY 9, 2011
78 oral
80 oral
HIGH-DOSE RATE BRACHYTHERAPY AS AN AMBULATORY BOOST TECHNIQUE IN BREAST CANCER : EXPERIENCE IN 502 PATIENTS S. Rivera1 , N. Taright1 , L. Quero1 , C. Maylin1 , C. Hennequin1 1 H ÔPITAL S AINT-L OUIS, Department of Radiation Oncology, Paris, France
LOCALISATION OF THE SURGICAL CAVITY USING SUPINE MAGNETIC RESONANCE AND COMPUTED TOMOGRAPHY SCAN FUSION FOR PLANIFICATION OF BREAST INTERSTITIAL BRACHYTHERAPY M. Jolicoeur1 , M. L. Racine1 , L. Hathout2 , S. David3 , I. Trop4 1 H OPITAL C HARLES L EMOYNE, Radiation Oncology, Longueuil, Canada 2 CHUM - C AMPUS N OTRE -DAME, radiation-oncology, Montréal, Canada 3 CHUM - C AMPUS N OTRE -DAME, Department of Medical Physics and Radiation Safety, Montréal, Canada 4 CHUM - C AMPUS N OTRE -DAME, Radiology, Montréal, Canada
Purpose: A boost dose is mandatory in most of the patients following irradiation of the mammary gland after lumpectomy. The best way to deliver this boost dose is unknown: simplicity and efficiency are required. HDR brachytherapy could be performed in one day under local anaesthesia. Materials: Implant under local anaesthesia is performed early in the morning, 2 to 3 weeks after external beam irradiation (EBRT: 44 Gy). Verification films and dosimetry (according to Paris’ system) are then immediately done, and a first dose of 5 Gy is delivered usually 90 to 120 minutes after the beginning of the implant. A second dose of 5 Gy is delivered 4 to 6 hours later in the afternoon. The patient is going back home at the end of the day.Patients: 502 patients were treated between 1990 and 2003. Mean age was 58. Mean histological size was 15 mm; most of the patients (80%) had an infiltrating ductal carcinoma, the others having a lobular infiltrating carcinoma or a DCIS. Surgical margins were negative in 66% of the cases, less than 2 mm in 24% and positive in 4.8%. Results: With a median follow-up of 5.5 years, 5-year local relapse rate is 4.8%, most of the relapse occurring in the same area. Most of the relapse occurs in the first four year of follow-up. The only prognostic factor of local relapse in multivariate analysis was young age (≤ 40 years). Pathological status of margins was not significantly associated with local control. Half of the patients had a cosmetic evaluation with a satisfactory result in more than 80% of the cases. The main adverse effect was fibrosis in the brachytherapy area. Conclusions: In view of our results, HDR brachytherapy could be proposed as an alternative to EBRT to boost the tumour bed. It could be performed in one day under local anaesthesia in experienced hands. 79 oral A SAFE AND PREDICTABLE METHOD OF DELIVERING THERMAL BOOST COMBINED WITH HDR BRACHYTHERAPY IN CONSERVATIVE MANAGEMENT OF BREAST CANCER A. Chichel1 , J. Skowronek1 , M. Kanikowski1 1 G REATER P OLAND C ANCER C ENTRE, Brachytherapy Department, Poznan, Poland Purpose: Standard breast conserving therapy (BCT) often uses interstitial multi-catheter HDR brachytherapy (BT) to minimize the risk of local relapse within the treated breast. Contemporary HDR BT offers conformal (CT based), accurate and effective irradiation of the tumor bed. Nevertheless, there is still some place for further improvement in local recurrence rate. Additional local hyperthermia (HT), substantially used as an adjuvant therapy, is probably one of the most effective radiation sensitizers known. HT (thermal boost) preceding standard BT boost has a potential of increasing the probability of local cure. The study is to report a new method of enhancing HDR brachytherapy boost by interstitial microwave hyperthermia (MV HT, thermal boost) in conservative management (BCT) of early breast cancer. Materials: Between February 2006 and December 2007, 57 stage IA-IIIA breast cancer patients were treated with BCT ± adjuvant chemotherapy followed by 42,5 50 Gy whole breast irradiation (WBI) and 10 Gy HDR BT boost. 32 patients (56,1 %) were treated with additional pre-BT single session of interstitial MW HT to a tumor bed. Tiny microwave antennas were inserted into the treated volume via brachytherapy applicators (multi-catheter technique, same implant for both treatments). Sets of planning CT scans provided strict geometrical information about 3D orientation of applicators within the tumor bed and enabled planning of the best pattern of MW antennas and thermometers to safely and efficiently heat the target volume before irradiation. Such approach enabled saving the skin and pectoral muscle form heating. Reference temperature was set on 43 C and intended therapeutic time (TT) was 1 hour. In addition, the incidence, severity and duration of early side effects in group with (I) or without hyperthermia (II) were assessed; significant p-value ≤ 0,05. Results: All 32 patients tolerated HT very well. HT sessions met the intended requirements (median): reference temperature 42,2 C, therapeutic time (TT) 61,4 minutes, total thermal dose 42 minutes and a gap between HT and BT 30 minutes. Median follow-up was 40 months with no local recurrences. Grade I and II radiodermatitis occurred in 24 (42,1 %) and 6 (10,5 %) patients, respectively, no significant difference between group I and II. Skin oedema and erythema occurred in 48 (85,7 %) and 36 (64,3 %) cases, respectively, both were equally distributed between the groups. Oedema was present up to 12 months and pattern of its regression was similar in group I and II, p = 0,933. Conclusions: Combination of HT and standard HDR BT boost was feasible and well tolerated. Pre-BT HT, having potential of improvement in breast cancer loca, didn’t increase early toxicity in treated patients. All side effects were transient and remained for up tol control 12 months. The study has to be continued on a larger and randomized group of patients to verify the potential of local control improvement and to assess the risk and profile of late toxicity.
Purpose: To evaluate the feasibility of a supine breast magnetic resonance imaging (MRI) scanning for definition and localisation of the surgical bed after breast conservative surgery Materials: Patients were eligible for this study if they had conservative surgery for breast cancer were candidates for brachytherapy as a component of their breast conservative treatment and had no contra-indications for MRI. To ensure position reproducibility, patients were placed supine position with ipsilateral arm above the head in an immobilization device. Localization markers were placed around the nipple, on the sternal notch and on the skin along the scar. All patients underwent planification CT and MRI exams in the same implant/treatment position. The exam consisted in acquisitions of axial and sagittal T1 and T2 images at 3-mm intervals. 4 points were predefined for CT-MRI image fusion. The surgical cavity was drawn on CT, then MRI, separately by three independent observers. The MRI was a Siemens 1.5 Tesla with a body coil. Fusion and analysis of CT and MRI images was performed using the ECLIPSE treatment planning software. The surgical cavity was drawn on CT, then MRI, separately by three independent observers. Results: From September 2005 to November 2008, 70 patients were included in this prospective study. For each patients, we were able to acquire axial T1 and T2 images of good quality. For image fusion, axial T2 images were preferred. Using the predefined points the mean error following the fusion was 5.57 mm. Volumes obtained on MRI were respectively 30%, 38% and 40% smaller than those determined from CT images for each observers. There was a good correlation between the localization of the surgical bed on MRI and CT with 85-95% of the MRI volume contained within the CT volume. Inter-reader variability in the estimation of surgical bed volume was 10-32 % on both exams. Conclusions: Supine breast MRI yields a more precise definition of the surgical bed than CT and obviate the need for surgical clips. The volume of the cavity is smaller with MRI. We propose that CT-MRI fusion to be use in the delineation of the surgical cavity in view of partial breast irradiation 81 oral BREAST TISSUE MODELING IN MONTE CARLO POST-IMPLANT EVALUATION OF BREAST LDR BRACHYTHERAPY H. Afsharpour1 , G. Landry2 , B. Reniers3 , J. P. Pignol4 , L. Beaulieu1 , F. Verhaegen5 1 C ENTRE H OSPITALIER U NIVERSITAIRE DE Q UÉBEC, L’H ÔTEL -D IEU DE Q UÉBEC, Quebec, Canada 2 M AASTRICHT R ADIATION O NCOLOGY (MAASTRO CLINIC ), Maastricht, Netherlands 3 MAASTRO CLINIC, Maastricht, Netherlands 4 O DETTE C ANCER C ENTRE , S UNNYBROOK H EALTH S CIENCES C ENTRE, North York, Toronto, Canada 5 M AASTRICHT R ADIATION T HERAPY AND O NCOLOGY (MAASTRO) CLINIC, Maastricht, Netherlands Purpose: Breast tissue is mainly composed of glandular and adipose tissues. The proportion of glandular/adipose tissue varies between patients depending on different factors such as the age of the patient. In the special case of 103Pd LDR breast brachytherapy where low energy photons (21 keV) are involved, the dose distribution is highly sensitive to chemical composition of tissues. Different approaches have been proposed to model the breast tissue for Monte Carlo dose calculations. The aim of this work is to use several approaches to model the breast tissue and to compare their effect on postimplant dosimetry for a patient cohort treated with LDR brachytherapy. Materials: Two months post-implant CT images of 35 breast LDR patients along with the contours of different organs are used to recreate accurate patient geometries in a Geant4.9.3-based Monte Carlo dosimetry platform. Different approaches are used to model the breast tissue. The first approach is based on the TG43 recommendations for which all tissues are replaced by the unit-density water. The second approach uses water to create each voxel with its density assigned from the CT image. The third approach is to create the breast tissue with a spatially uniform mixture of glandular/adipose with various proportions determined from the mean breast density. A fourth approach uses a similar method but applies it voxel by voxel. Finally, a semi-automatic segmentation approach is proposed using a density threshold distinguishing the glandular from the adipose tissues inside the breast. To investigate the clinical impact, the D90 metric was extracted and compared between different formalisms. It is important to note that the radiation transport is performed in
M ONDAY, M AY 9, 2011
P ROFFERED PAPER
S 31
the medium while the dose is scored in water. Results: The 50-50 uniform breast model represents the highest dose to the PTV and the lowest dose to the skin compared to the TG43 formalism, which is predicting the lowest doses to the PTV. Looking at the DVHs for one patient (see the figure below), the D90 value is about 110 Gy in the 50-50 uniform approach but it decreases to 98 Gy when the TG43 formalism is used. Using the second approach (water with varying density) the same patient has a D90 value of about 97 Gy which is very close to what is predicted by the TG43 formalism. All tissue assignment approaches will be compared for 35 patients.
Conclusions: External validation of the nomogram shows adequate discrimination of AUR, but underestimation of the actual AUR risk. Since the nomogram is able to correctly identify low risk patients with a negative predictive value of 98%, the nomogram can aid in individualized treatment decisionmaking.
Conclusions: Different tissue modeling schemes were used and compared in this study. The resulting dose distributions and clinical metrics are quite different when comparing these formalisms. This argues in favor of an agreement for a standard tissue modeling approach to be used everywhere in the clinics. Semiautomatic segmentation seems promising and feasible for more accurate dosimetry in LDR breast brachytherapy.
GEC Proffered Papers 2: Prostate Toxicity 82 oral EXTERNAL VALIDATION OF THE PRE-TREATMENT NOMOGRAM TO PREDICT ACUTE URINARY RETENTION AFTER I-125 PROSTATE BRACHYTHERAPY M. van Vulpen1 , E. Roeloffzen1 , E. Monninkhof2 , S. Elantholi Parameswaran3 U.M.C. U TRECHT, Radiation Oncology, Utrecht, Netherlands 2 U.M.C. U TRECHT, Epidemiology, Utrecht, Netherlands 3 P RINCESS M ARGARET H OSPITAL, Radiation Oncology, Toronto, Canada
1
Purpose: Acute urinary retention (AUR) after Iodine-125 (I-125) prostate brachytherapy has a negative impact on quality of life. Recently, we developed a nomogram to preoperatively predict the risk of AUR. The aim of this study was to assess the external validity of the nomogram. Materials: The nomogram was initially developed on 714 patients treated with I-125 prostate brachytherapy at the University Medical Center Utrecht, The Netherlands. Predictive factors included in the nomogram were prostate volume, IPSS, neo-adjuvant hormonal treatment and prostate protrusion. For external validation, the data of 715 consecutive patients who underwent I-125 prostate brachytherapy between January 2003 and July 2008 at the Princess Margaret Hospital (PMH), Toronto were used. The performance of the nomogram was evaluated by discrimination (ability to distinguish between patients who develop AUR yes or no) and calibration (agreement between observed and predicted numbers of AUR). Results: Of the 715 patients treated at the PMH, 67 patients (9.4%) developed AUR, compared to 8.0% in the UMCU cohort. In the validation dataset, the discriminatory ability of the nomogram was good (ROC area 0.86; 95% CI 0.82 0.91), and comparable to the derivation dataset (ROC area 0.82; 95% CI 0.77 0.88). Comparison between predicted risks and observed frequencies of AUR showed systematic underestimation of the nomogram in the validation dataset. Still, the negative predictive value for the risk of AUR, using a cut-off value of 5%, was high (98.1%).
83 oral URETHRA-SPARING, INTRA-OPERATIVE, REAL-TIME PLANNED, PERMANENT-SEED PROSTATE BRACHYTHERAPY: TOXICITY ANALYSIS T. Zilli1 , D. Taussky2 , D. Donath2 , P. L. Hoa3 , R. X. Larouche3 , D. Béliveau-Nadeau3 , Y. Hervieux3 , G. Delouya3 1 C ENTRE HOSPITALIER DE L’U NIVERSITÉ DE M ONTRÉAL, Radiation Oncology, Montreal, Canada 2 CHUM - C AMPUS N OTRE -DAME, Montréal, Canada 3 CHUM - C AMPUS N OTRE -DAME, Radiation Oncology, Montréal, Canada Purpose: To report the toxicity outcome in patients with localized prostate cancer undergoing 125 I permanent-seed brachytherapy (BT) according to an urethra-sparing, intra-operative (IO), real-time planned conformal technique. Materials: Data were analyzed on 250 patients treated consecutively for lowor intermediate-risk prostate cancer between 2005 and 2009. The planned goal was urethral V150 =0%. Acute and late genito-urinary (GU), gastrointestinal (GI) and erectile toxicities were scored with the International Prostate Symptom Score (IPSS) questionnaire and Common Terminology Criteria for Adverse Events (version 3.0). Median follow-up time for patients with at least 2 years of follow-up (n=130) was 34.4 months (range 24-56.9). Results: Mean IO urethral V150, D5 and D2 were 0.018±0.08%, 193±24.3 Gy and 199±16.3 Gy, respectively. Mean prostate D90 and V100 on day30 computed tomography scan were 158.0±27.0 Gy and 92.1±7.2%, respectively. Low-activity sources (median activity 0.44 mCi, range 0.39-0.49) were implanted in 113 (45.2%) patients, whereas 137 (54.8%) patients were implanted with high-activity sources (median activity 0.61 mCi, range 0.540.68). Mean IPSS peak was 9.5±6.3 1 month after BT (mean difference from baseline IPSS=5.3). The mean difference between baseline and postimplantation IPSS was 5.3 at 1 month, 4.6 at 4 months, 2.1 at 8 months, and 0.5 at 12 months. 96.8% of the patients were receiving alpha-blockers 1 month after implantation, and this percentage decreased to 84.1%, 72.3% and 57.3% at 4, 8 and 12 months, respectively. No acute GI toxicity was observed in 86.8% of patients. Overall, 31 (12.4%) and 2 patients presented with acute Grade 1 and Grade 2 GI toxicity, respectively. The 3-year probability of Grade ≥2 late GU toxicity-free survival was 77.4±4.0%, with Grade 3 late GU toxicity encountered in only 3 patients( 3-year probability rate of Grade ≥3 late GU toxicity-free survival of 97.5±1.9%) . Three-year Grade 1 late GI toxicity-free survival was 86.1±3.2%. No patient presented Grade ≥2 late GI toxicity. At last follow-up, the rate of patients without erectile dysfunction decreased from 59.2% to 30.8%. 20.7% of patients with normal sexual status at baseline manifested Grade ≥2 erectile dysfunction after BT. On multivariate analysis, elevated baseline IPSS (p=0.016) and high-activity sources (median 0.61 mCi) (p=0.033) predicted increased Grade ≥2 late GU toxicity. Conclusions: Urethra-sparing IO BT is feasible and results in low acute and late GU toxicity compared to the literature. High seed activity seems to increase the rate of mild to severe long-term urinary toxicity.
P ROFFERED PAPER
M ONDAY, M AY 9, 2011
78 oral
80 oral
HIGH-DOSE RATE BRACHYTHERAPY AS AN AMBULATORY BOOST TECHNIQUE IN BREAST CANCER : EXPERIENCE IN 502 PATIENTS S. Rivera1 , N. Taright1 , L. Quero1 , C. Maylin1 , C. Hennequin1 1 H ÔPITAL S AINT-L OUIS, Department of Radiation Oncology, Paris, France
LOCALISATION OF THE SURGICAL CAVITY USING SUPINE MAGNETIC RESONANCE AND COMPUTED TOMOGRAPHY SCAN FUSION FOR PLANIFICATION OF BREAST INTERSTITIAL BRACHYTHERAPY M. Jolicoeur1 , M. L. Racine1 , L. Hathout2 , S. David3 , I. Trop4 1 H OPITAL C HARLES L EMOYNE, Radiation Oncology, Longueuil, Canada 2 CHUM - C AMPUS N OTRE -DAME, radiation-oncology, Montréal, Canada 3 CHUM - C AMPUS N OTRE -DAME, Department of Medical Physics and Radiation Safety, Montréal, Canada 4 CHUM - C AMPUS N OTRE -DAME, Radiology, Montréal, Canada
Purpose: A boost dose is mandatory in most of the patients following irradiation of the mammary gland after lumpectomy. The best way to deliver this boost dose is unknown: simplicity and efficiency are required. HDR brachytherapy could be performed in one day under local anaesthesia. Materials: Implant under local anaesthesia is performed early in the morning, 2 to 3 weeks after external beam irradiation (EBRT: 44 Gy). Verification films and dosimetry (according to Paris’ system) are then immediately done, and a first dose of 5 Gy is delivered usually 90 to 120 minutes after the beginning of the implant. A second dose of 5 Gy is delivered 4 to 6 hours later in the afternoon. The patient is going back home at the end of the day.Patients: 502 patients were treated between 1990 and 2003. Mean age was 58. Mean histological size was 15 mm; most of the patients (80%) had an infiltrating ductal carcinoma, the others having a lobular infiltrating carcinoma or a DCIS. Surgical margins were negative in 66% of the cases, less than 2 mm in 24% and positive in 4.8%. Results: With a median follow-up of 5.5 years, 5-year local relapse rate is 4.8%, most of the relapse occurring in the same area. Most of the relapse occurs in the first four year of follow-up. The only prognostic factor of local relapse in multivariate analysis was young age (≤ 40 years). Pathological status of margins was not significantly associated with local control. Half of the patients had a cosmetic evaluation with a satisfactory result in more than 80% of the cases. The main adverse effect was fibrosis in the brachytherapy area. Conclusions: In view of our results, HDR brachytherapy could be proposed as an alternative to EBRT to boost the tumour bed. It could be performed in one day under local anaesthesia in experienced hands. 79 oral A SAFE AND PREDICTABLE METHOD OF DELIVERING THERMAL BOOST COMBINED WITH HDR BRACHYTHERAPY IN CONSERVATIVE MANAGEMENT OF BREAST CANCER A. Chichel1 , J. Skowronek1 , M. Kanikowski1 1 G REATER P OLAND C ANCER C ENTRE, Brachytherapy Department, Poznan, Poland Purpose: Standard breast conserving therapy (BCT) often uses interstitial multi-catheter HDR brachytherapy (BT) to minimize the risk of local relapse within the treated breast. Contemporary HDR BT offers conformal (CT based), accurate and effective irradiation of the tumor bed. Nevertheless, there is still some place for further improvement in local recurrence rate. Additional local hyperthermia (HT), substantially used as an adjuvant therapy, is probably one of the most effective radiation sensitizers known. HT (thermal boost) preceding standard BT boost has a potential of increasing the probability of local cure. The study is to report a new method of enhancing HDR brachytherapy boost by interstitial microwave hyperthermia (MV HT, thermal boost) in conservative management (BCT) of early breast cancer. Materials: Between February 2006 and December 2007, 57 stage IA-IIIA breast cancer patients were treated with BCT ± adjuvant chemotherapy followed by 42,5 50 Gy whole breast irradiation (WBI) and 10 Gy HDR BT boost. 32 patients (56,1 %) were treated with additional pre-BT single session of interstitial MW HT to a tumor bed. Tiny microwave antennas were inserted into the treated volume via brachytherapy applicators (multi-catheter technique, same implant for both treatments). Sets of planning CT scans provided strict geometrical information about 3D orientation of applicators within the tumor bed and enabled planning of the best pattern of MW antennas and thermometers to safely and efficiently heat the target volume before irradiation. Such approach enabled saving the skin and pectoral muscle form heating. Reference temperature was set on 43 C and intended therapeutic time (TT) was 1 hour. In addition, the incidence, severity and duration of early side effects in group with (I) or without hyperthermia (II) were assessed; significant p-value ≤ 0,05. Results: All 32 patients tolerated HT very well. HT sessions met the intended requirements (median): reference temperature 42,2 C, therapeutic time (TT) 61,4 minutes, total thermal dose 42 minutes and a gap between HT and BT 30 minutes. Median follow-up was 40 months with no local recurrences. Grade I and II radiodermatitis occurred in 24 (42,1 %) and 6 (10,5 %) patients, respectively, no significant difference between group I and II. Skin oedema and erythema occurred in 48 (85,7 %) and 36 (64,3 %) cases, respectively, both were equally distributed between the groups. Oedema was present up to 12 months and pattern of its regression was similar in group I and II, p = 0,933. Conclusions: Combination of HT and standard HDR BT boost was feasible and well tolerated. Pre-BT HT, having potential of improvement in breast cancer loca, didn’t increase early toxicity in treated patients. All side effects were transient and remained for up tol control 12 months. The study has to be continued on a larger and randomized group of patients to verify the potential of local control improvement and to assess the risk and profile of late toxicity.
Purpose: To evaluate the feasibility of a supine breast magnetic resonance imaging (MRI) scanning for definition and localisation of the surgical bed after breast conservative surgery Materials: Patients were eligible for this study if they had conservative surgery for breast cancer were candidates for brachytherapy as a component of their breast conservative treatment and had no contra-indications for MRI. To ensure position reproducibility, patients were placed supine position with ipsilateral arm above the head in an immobilization device. Localization markers were placed around the nipple, on the sternal notch and on the skin along the scar. All patients underwent planification CT and MRI exams in the same implant/treatment position. The exam consisted in acquisitions of axial and sagittal T1 and T2 images at 3-mm intervals. 4 points were predefined for CT-MRI image fusion. The surgical cavity was drawn on CT, then MRI, separately by three independent observers. The MRI was a Siemens 1.5 Tesla with a body coil. Fusion and analysis of CT and MRI images was performed using the ECLIPSE treatment planning software. The surgical cavity was drawn on CT, then MRI, separately by three independent observers. Results: From September 2005 to November 2008, 70 patients were included in this prospective study. For each patients, we were able to acquire axial T1 and T2 images of good quality. For image fusion, axial T2 images were preferred. Using the predefined points the mean error following the fusion was 5.57 mm. Volumes obtained on MRI were respectively 30%, 38% and 40% smaller than those determined from CT images for each observers. There was a good correlation between the localization of the surgical bed on MRI and CT with 85-95% of the MRI volume contained within the CT volume. Inter-reader variability in the estimation of surgical bed volume was 10-32 % on both exams. Conclusions: Supine breast MRI yields a more precise definition of the surgical bed than CT and obviate the need for surgical clips. The volume of the cavity is smaller with MRI. We propose that CT-MRI fusion to be use in the delineation of the surgical cavity in view of partial breast irradiation 81 oral BREAST TISSUE MODELING IN MONTE CARLO POST-IMPLANT EVALUATION OF BREAST LDR BRACHYTHERAPY H. Afsharpour1 , G. Landry2 , B. Reniers3 , J. P. Pignol4 , L. Beaulieu1 , F. Verhaegen5 1 C ENTRE H OSPITALIER U NIVERSITAIRE DE Q UÉBEC, L’H ÔTEL -D IEU DE Q UÉBEC, Quebec, Canada 2 M AASTRICHT R ADIATION O NCOLOGY (MAASTRO CLINIC ), Maastricht, Netherlands 3 MAASTRO CLINIC, Maastricht, Netherlands 4 O DETTE C ANCER C ENTRE , S UNNYBROOK H EALTH S CIENCES C ENTRE, North York, Toronto, Canada 5 M AASTRICHT R ADIATION T HERAPY AND O NCOLOGY (MAASTRO) CLINIC, Maastricht, Netherlands Purpose: Breast tissue is mainly composed of glandular and adipose tissues. The proportion of glandular/adipose tissue varies between patients depending on different factors such as the age of the patient. In the special case of 103Pd LDR breast brachytherapy where low energy photons (21 keV) are involved, the dose distribution is highly sensitive to chemical composition of tissues. Different approaches have been proposed to model the breast tissue for Monte Carlo dose calculations. The aim of this work is to use several approaches to model the breast tissue and to compare their effect on postimplant dosimetry for a patient cohort treated with LDR brachytherapy. Materials: Two months post-implant CT images of 35 breast LDR patients along with the contours of different organs are used to recreate accurate patient geometries in a Geant4.9.3-based Monte Carlo dosimetry platform. Different approaches are used to model the breast tissue. The first approach is based on the TG43 recommendations for which all tissues are replaced by the unit-density water. The second approach uses water to create each voxel with its density assigned from the CT image. The third approach is to create the breast tissue with a spatially uniform mixture of glandular/adipose with various proportions determined from the mean breast density. A fourth approach uses a similar method but applies it voxel by voxel. Finally, a semi-automatic segmentation approach is proposed using a density threshold distinguishing the glandular from the adipose tissues inside the breast. To investigate the clinical impact, the D90 metric was extracted and compared between different formalisms. It is important to note that the radiation transport is performed in
M ONDAY, M AY 9, 2011
P ROFFERED PAPER
S 31
the medium while the dose is scored in water. Results: The 50-50 uniform breast model represents the highest dose to the PTV and the lowest dose to the skin compared to the TG43 formalism, which is predicting the lowest doses to the PTV. Looking at the DVHs for one patient (see the figure below), the D90 value is about 110 Gy in the 50-50 uniform approach but it decreases to 98 Gy when the TG43 formalism is used. Using the second approach (water with varying density) the same patient has a D90 value of about 97 Gy which is very close to what is predicted by the TG43 formalism. All tissue assignment approaches will be compared for 35 patients.
Conclusions: External validation of the nomogram shows adequate discrimination of AUR, but underestimation of the actual AUR risk. Since the nomogram is able to correctly identify low risk patients with a negative predictive value of 98%, the nomogram can aid in individualized treatment decisionmaking.
Conclusions: Different tissue modeling schemes were used and compared in this study. The resulting dose distributions and clinical metrics are quite different when comparing these formalisms. This argues in favor of an agreement for a standard tissue modeling approach to be used everywhere in the clinics. Semiautomatic segmentation seems promising and feasible for more accurate dosimetry in LDR breast brachytherapy.
GEC Proffered Papers 2: Prostate Toxicity 82 oral EXTERNAL VALIDATION OF THE PRE-TREATMENT NOMOGRAM TO PREDICT ACUTE URINARY RETENTION AFTER I-125 PROSTATE BRACHYTHERAPY M. van Vulpen1 , E. Roeloffzen1 , E. Monninkhof2 , S. Elantholi Parameswaran3 U.M.C. U TRECHT, Radiation Oncology, Utrecht, Netherlands 2 U.M.C. U TRECHT, Epidemiology, Utrecht, Netherlands 3 P RINCESS M ARGARET H OSPITAL, Radiation Oncology, Toronto, Canada
1
Purpose: Acute urinary retention (AUR) after Iodine-125 (I-125) prostate brachytherapy has a negative impact on quality of life. Recently, we developed a nomogram to preoperatively predict the risk of AUR. The aim of this study was to assess the external validity of the nomogram. Materials: The nomogram was initially developed on 714 patients treated with I-125 prostate brachytherapy at the University Medical Center Utrecht, The Netherlands. Predictive factors included in the nomogram were prostate volume, IPSS, neo-adjuvant hormonal treatment and prostate protrusion. For external validation, the data of 715 consecutive patients who underwent I-125 prostate brachytherapy between January 2003 and July 2008 at the Princess Margaret Hospital (PMH), Toronto were used. The performance of the nomogram was evaluated by discrimination (ability to distinguish between patients who develop AUR yes or no) and calibration (agreement between observed and predicted numbers of AUR). Results: Of the 715 patients treated at the PMH, 67 patients (9.4%) developed AUR, compared to 8.0% in the UMCU cohort. In the validation dataset, the discriminatory ability of the nomogram was good (ROC area 0.86; 95% CI 0.82 0.91), and comparable to the derivation dataset (ROC area 0.82; 95% CI 0.77 0.88). Comparison between predicted risks and observed frequencies of AUR showed systematic underestimation of the nomogram in the validation dataset. Still, the negative predictive value for the risk of AUR, using a cut-off value of 5%, was high (98.1%).
83 oral URETHRA-SPARING, INTRA-OPERATIVE, REAL-TIME PLANNED, PERMANENT-SEED PROSTATE BRACHYTHERAPY: TOXICITY ANALYSIS T. Zilli1 , D. Taussky2 , D. Donath2 , P. L. Hoa3 , R. X. Larouche3 , D. Béliveau-Nadeau3 , Y. Hervieux3 , G. Delouya3 1 C ENTRE HOSPITALIER DE L’U NIVERSITÉ DE M ONTRÉAL, Radiation Oncology, Montreal, Canada 2 CHUM - C AMPUS N OTRE -DAME, Montréal, Canada 3 CHUM - C AMPUS N OTRE -DAME, Radiation Oncology, Montréal, Canada Purpose: To report the toxicity outcome in patients with localized prostate cancer undergoing 125 I permanent-seed brachytherapy (BT) according to an urethra-sparing, intra-operative (IO), real-time planned conformal technique. Materials: Data were analyzed on 250 patients treated consecutively for lowor intermediate-risk prostate cancer between 2005 and 2009. The planned goal was urethral V150 =0%. Acute and late genito-urinary (GU), gastrointestinal (GI) and erectile toxicities were scored with the International Prostate Symptom Score (IPSS) questionnaire and Common Terminology Criteria for Adverse Events (version 3.0). Median follow-up time for patients with at least 2 years of follow-up (n=130) was 34.4 months (range 24-56.9). Results: Mean IO urethral V150, D5 and D2 were 0.018±0.08%, 193±24.3 Gy and 199±16.3 Gy, respectively. Mean prostate D90 and V100 on day30 computed tomography scan were 158.0±27.0 Gy and 92.1±7.2%, respectively. Low-activity sources (median activity 0.44 mCi, range 0.39-0.49) were implanted in 113 (45.2%) patients, whereas 137 (54.8%) patients were implanted with high-activity sources (median activity 0.61 mCi, range 0.540.68). Mean IPSS peak was 9.5±6.3 1 month after BT (mean difference from baseline IPSS=5.3). The mean difference between baseline and postimplantation IPSS was 5.3 at 1 month, 4.6 at 4 months, 2.1 at 8 months, and 0.5 at 12 months. 96.8% of the patients were receiving alpha-blockers 1 month after implantation, and this percentage decreased to 84.1%, 72.3% and 57.3% at 4, 8 and 12 months, respectively. No acute GI toxicity was observed in 86.8% of patients. Overall, 31 (12.4%) and 2 patients presented with acute Grade 1 and Grade 2 GI toxicity, respectively. The 3-year probability of Grade ≥2 late GU toxicity-free survival was 77.4±4.0%, with Grade 3 late GU toxicity encountered in only 3 patients( 3-year probability rate of Grade ≥3 late GU toxicity-free survival of 97.5±1.9%) . Three-year Grade 1 late GI toxicity-free survival was 86.1±3.2%. No patient presented Grade ≥2 late GI toxicity. At last follow-up, the rate of patients without erectile dysfunction decreased from 59.2% to 30.8%. 20.7% of patients with normal sexual status at baseline manifested Grade ≥2 erectile dysfunction after BT. On multivariate analysis, elevated baseline IPSS (p=0.016) and high-activity sources (median 0.61 mCi) (p=0.033) predicted increased Grade ≥2 late GU toxicity. Conclusions: Urethra-sparing IO BT is feasible and results in low acute and late GU toxicity compared to the literature. High seed activity seems to increase the rate of mild to severe long-term urinary toxicity.