- Email: [email protected]
A case report on bilateral partial breast irradiation using SAVI
Medical Dosimetry 37 (2012) 214-220
Medical Dosimetry journal homepage: www.meddos.org
A case report on bilateral partial breast irradiation using SAVI Aime M. Gloi, Ph.D.,* Robert Buchanan M.Sc., D.A.B.R.,† Jeff Nuskind, M.S., D.A.B.R.,* Corrie Zuge, B.S., C.M.D.,* and Anndrea Goettler, C.M.D.* *Radiation Oncology, St. Vincent Hospital, Green Bay, WI; and †Southeast Alabama Medical Center, Radiation Oncology Department, Dothan, AL 36301
A R T I C L E
I N F O
Article history: Received 2 May 2011 Accepted 10 August 2011 Keywords: SAVI DVH DHI ODI DNR EUD NTCP TCP
A B S T R A C T To assess dosimetric parameters in a case study where bilateral accelerated partial breast irradiation (APBI) is delivered using a strut-adjusted volume implant (SAVI) device. A 59-year-old female received APBI in both breasts over 5 days, with fractions of 3.4 Gy twice daily. A Vac-lok system was used for immobilization, and a C-arm was used for daily imaging. We generated dose-volume histograms (DVHs) for the brachytherapy plans to derive several important biologic factors. We calculated the normal tissue complication probability (NTCP), equivalent uniform dose (EUD), and tumor control probability (TCP) using the Lyman-Kutcher-Burman model parameters ␣ ⫽ 0.3 Gy⫺1, ␣/ ⫽ 4 Gy, n ⫽ 0.1, and m ⫽ 0.3. In addition, we assessed the dose homogeneity index (DHI), overdose index, and dose nonuniformity ratio. D95 was ⬎95% and V150 was ⬍50 mL for both breasts. The DHIs were 0.469 and 0.512 for the left and right breasts, respectively. The EUDs (normalized to 3.4 Gy b.i.d.) were 33.53 and 29.10 Gy. The TCPs were estimated at 99.2% and 99.9%, whereas the NTCP values were 4.2% and 2.57%. In this clinical case, we were able to quantify the dosimetric parameters of an APBI treatment performed with a SAVI device. 䉷 2012 American Association of Medical Dosimetrists.
Introduction Breast cancer is one of the most common types of cancer among women in the United States. According to a study by Fewer et al., on average 192,370 women are diagnosed with the disease every year, and approximately 20% of those women will die of it.1 Early-stage breast cancer is characterized by a lack of lymph nodes, metastasis, and clinical lesions with diameters up to 2 cm. In recent years, high public awareness and early mammography screening have increased the number of women diagnosed with early-stage breast cancer. The treatment method of choice for many of these women (40%) is a lumpectomy followed by accelerated partial breast irradiation (APBI).1 This treatment has 2 major advantages compared with other available methods of removing the cancerous tissue: it requires a shorter time commitment (5 days vs several weeks) and it is less invasive. An abundance of devices have been developed for APBI treatment. For example, the Mammosite RTS (Cytyc, Corp., Marlborough, MA) consists of a balloon with a single central lumen. The geometry of the balloon permits the delivery of an increased dose because the radiation is limited to tissue within 1 cm of the balloon’s surface. Good to excellent cosmetic results have been achieved with a MammoSite
catheter when the balloon-to-skin spacing is ⬎7 mm. Skin necrosis and poor cosmesis are observed in several cases when the spacing is ⬍7 mm2. Not all patients are eligible for the device; MammoSite presents some stringent criteria, such as tissue conformance to the balloon, breast size, and other geometrical factors. Following the MammoSite example, the multilumen balloon Contura (SenoRx, Inc., Aliso Viejo, CA) was developed to minimize the radiation “hot spot” in the skin and reduce air/seroma volume. Contura allows for more flexible dosimetry and adapts more easily to geometrical constraints, such as the skin’s distance from the catheter. A hybrid device named SAVI (Strut-Adjusted Volume Implant) from (Cianna Medical, Aliso Viejo, CA), was created to eliminate or minimize the deficiencies of MammoSite and Contura. SAVI provides even more customized radiation therapy, improving the dosimetry and considerably reducing the dose to the skin and chest wall. SAVI also provides more options and precision in its planning system, allowing physicians to tailor the radiation to the tumor site.3 The aim of this study was to analyze the dosimetric parameters of an unusual clinical case, in which the patient received bilateral APBI treatment via a SAVI device. Methods and Materials
Reprint requests to: Aime M. Gloi, Ph.D., Radiation Oncology, St. Vincent Hospital, 835 S. Van Buren Street, PO Box 13508, Green Bay, WI 54307-3508. E-mail: [email protected]
The subject of this report is a 59-year-old female with a breast abnormality noted on a mammogram screening. At the 6-month follow-up visit, the mammogram showed an increase in size and the magnetic resonance image revealed an enhanced area in the
0958-3947/$ – see front matter Copyright 䉷 2012 American Association of Medical Dosimetrists doi:10.1016/j.meddos.2011.08.001
A.M. Gloi et al. / Medical Dosimetry 37 (2012) 214-220
breast. A core biopsy was performed, showing poorly differentiated invasive ductal carcinoma with the globular features typically associated with ductal carcinoma in situ (DCIS). The disease was staged as T2, N0, and M0 grade 3 infiltrating ductal carcinoma of both breasts. The patient was implanted with a SAVI 6 –1 (6 peripheral catheters and one central catheter) device in the left breast, and a SAVI 8 –1 (8 peripheral and one central) in the right breast. The patient was immobilized in a Vac-lok (Medtech, Orange City, IA) cushion marked for reproducible positioning. These cushions are designed for comfortable, patient-specific repositioning; they form a custom mold of the patient’s anatomical contours in any desired pose. The patient’s position during the computed tomography (CT) scan can therefore be duplicated exactly during treatment.
215
The lumpectomy CT images were acquired with the patient immobilized in the treatment position. The slice thickness of the scan was 2.0 mm. The lumpectomy cavity was identified with the help of surgical clips. A fluoroscopy from a C-arm (General Electric OES 9800, Milwaukee, WI) was performed on the day of the CT scan to gather a baseline measurement. The Brachyvision (Varian Medical, Inc., Palo Alto, CA) treatment planning system was used for contouring and planning. Both the cavity and normal tissues, such as skin, ipsilateral lungs, and ribs, were contoured. The planning treatment volume (PTV) was determined by adding a layer 1.0 cm thick around the cavity. In the treatment planning system, we define the volume PTV_EVAL as the PTV minus the cavity. There were 2
Fig. 1. (a, b) SAVI 6-1 inserted in the left breast.
216
A.M. Gloi et al. / Medical Dosimetry 37 (2012) 214-220
Fig. 2. (a, b) SAVI 8-1 inserted in the right breast. treatment planning goals: to limit the skin dose ⬍100% of the prescribed dose and to limit the volume of air/fluid in the cavity during treatment to ⬍10.0% of the PTV. The fractions were separated by an interval of at least 6 hours for radiobiologic purposes. Each breast received 3.4 Gy per fraction, twice daily, for a total of 10 fractions per breast over 5 days. This report uses 3 quality indexes: the dose homogeneity index (DHI), the overdose volume index (ODI), and the dose nonuniformity ratio (DNR). DHI is defined as the volumes that receive a dose of at least 1.0 times the prescription dose (V100) and at least 1.5 times the reference dose (V150).4 Each volume is expressed as a fraction of the target volume PTV and DHI is the difference between them:
DHI ⫽ 共V100 ⫺ V150兲 ⁄ PTV ODI is the ratio of the volume receiving at least twice the prescribed dose to the volume receiving at least the prescribed dose4:
ODI ⫽ V200 ⁄ PTV Finally, we compute DNR as the ratio between the volume that receives at least 1.5 times the prescribed dose to the volume that receives at least the prescribed dose5:
DNR ⫽ V150 ⁄ PTV
A.M. Gloi et al. / Medical Dosimetry 37 (2012) 214-220
217
Fig. 3. Bilateral accelerated partial breast irradiation through SAVI.
We test the normal tissue complication probability (NTCP) model on the occurrence of breast fibrosis. The Lyman-Kutcher-Burman (LKB) model was used to calculate the NTCP. Based on the differential dose-volume histograms (DVHs), we calculate the equivalent uniform dose (EUD)6:
冉兺
EUD ⫽
i
vi vT
冊
共Di兲1⁄n
n
,
where vi is the fractional volume irradiated to dose Di, vT is the total volume of the structure, and n is a tissue-specific model fitting parameter. In this calculation, n is set to 0.1. Given the EUD, we evaluated the NTCP7 for a breast fibrosis as follows:
NTCP ⫽
1
兹2
兰
1
⫺⬁
2⁄2
e⫺x dx
where
t⫽
EUD ⫺ D50 m ⫻ D50
.
D50 is the prescribed dose with a 50% chance of complications, and m is another model parameter inversely proportional to the steepness of the dose response. Here, we use D50 ⫽ 70 Gy and m ⫽ 0.3.8, 9 The tumor control probability (TCP) relies on the fact that all cancer cells have been destroyed.8 It is given by
TCP ⫽ exp 共⫺K ⫻ S兲 , where k ⫽ 200 and the surviving fraction S is given by10
冋冉
S ⫽ exp ⫺ ␣ ⫻ EUD ⫹ d ⫻ EUD ⫺ 0.5␥
EUD d
冊册
.
To calculate S, we take the parameters of an external beam treatment with 2 Gy per fraction: ␣ ⫽ 0.3 Gy⫺1, ␣/ ⫽ 4 Gy, and ␥ ⫽ 0.693/Td (where Td is the tumor cell doubling time, or 15 d).11
Results The patient received a SAVI 6 –1 (Fig. 1a, 1b) in her left breast, and a SAVI 8 –1(Fig. 2a, 2b) in her right breast (Fig. 3). The mean cavity
volumes were 16.8 cm3 for the left breast and 30.4 cm3 for the right breast. The PTV_EVAL volume was greater in the right breast (70.2 cm3) than in the left breast (51.8 cm3). The DVHs are shown in Fig. 4a and 4b. V95, defined as the percentage of PTV_EVAL receiving at least 95% of the prescribed dose, was 95.75% and 98.76% for the left and right breasts, respectively. The skin dose in the left breast reached the limiting value of 100%. The skin dose in the right breast, however, was well below 100%. The DHI was 0.512 for the right breast and 0.469 for the left breast. The ODI was estimated at 0.177 for the left breast and 0.150 for the right breast. NTCP distributions were calculated for both breasts to assess the probability of breast fibrosis, using the values D50 ⫽ 70 Gy and m ⫽ 8, 9 0.3. The estimated probabilities were 4.27% for the left breast and 2.57% for the right breast. Notice that the left breast was implanted with a 6 –1 device, whereas an 8 –1 device was used for the right breast. The TCPs were 99.2% and 99.9% for the left and right breasts, respectively. Discussion All the dosimetric criteria for this APBI treatment were met for both breasts: V95 ⬎95%, V150 ⬍50 mL, and V200 ⬍20 mL. The data for this case compare well with the study of Gurdalli et al.,3 wherein 15 patients were treated with a variety of SAVI devices. Similar dose coverage was observed by Mancharan et al.12 The air/seroma volume was not a factor in this treatment, as is sometimes the case in MammoSite, having been contained by the struts. A SAVI (Fig. 5) device with several struts allows the doctor to shape the dose, resulting in good coverage. It is believed that all 3 devices (MammoSite, Contura, and SAVI) provide similar coverage of the target volume, but SAVI offers the most dose flexibility for patients whose skin and chest wall are close to the cavity.3
218
A.M. Gloi et al. / Medical Dosimetry 37 (2012) 214-220
Fig. 4. (a, b) DHH of right and left breast from top to bottom, respectively, describing the dose coverage of PTV and PTV_Eval.
A quantitative comparison of dose distributions in the 2 breasts is provided in Table 1. For an ideal implant, the quality indexes would be DHI ⫽ 1, ODI ⫽ 0, and DNR ⫽ 0. As anticipated, SAVI achieves a lower DHI than MammoSite: 0.59 (Gurdalli et al.3) in contrast to 0.66 (Yongbok et al.13) or 0.59 in contrast to 0.62 (Dickler et al.14). Because of the greater inhomogeneity of its radiation dose, MammoSite also treated a larger volume of breast tissue (between 110% and 170% of the prescribed dose) than did SAVI. In a recent paper, Gurdalli et al.3 reported results obtained with several SAVI devices. The mean DHI was 0.559 (range 0.52– 0.60),
which is somewhat higher than the indexes shown in Table 1 (0.469 and 0.512 for the left and right breasts, respectively). By contrast, Bloom et al.15 obtained a much lower mean DHI of 0.32. This inconsistent result shows that the methodology of APBI treatments requires more attention. Similar inconsistencies are observed in the DNR and ODI values. In our case, the low values of DNR and ODI suggest that the unnecessary irradiation of normal tissue around the target volume was negligible. Interestingly, Wazer et al.15 uncovered a correlation between DHI and cosmetic outcome in treatments using interstitial LDR boost breast implants.
A.M. Gloi et al. / Medical Dosimetry 37 (2012) 214-220
The available data are both sparse and noisy; consequently, any model predictions based on clinical data will have wide confidence ranges and could be considerably biased. Another problem with the evaluation of TCP and NTCP is the complexity of the radiobiological model and the difficulty of finding adequate fitting parameters. The LKB model certainly does not adequately reflect known radiobiological theory, because the radiation response can be highly dependent on the treatment scheme. In addition, the model assumes a homogeneous organ so it does take into account tissue damage at contiguous sites. Also, it is well understood that the breast DVH may not be consistent during CT and treatment because of variations in pose and the patient’s breathing motion. This problem may cause the overall accuracy to suffer, even if a C-arm is used for daily imaging. In this report, the TCP for posttreatment fibrosis was calculated through the EUD for a nonuniform dose. Our predictions are comparable with those of several authors.3,5,10,16-18 The NTCP and TCP can be calculated using treatment-only variables and the radiobiological characteristics of tissue. Despite the limitations of the model, this case study has provided useful insight into the radiobiological parameters of normal tissue responses.
Fig. 5. SAVI device.
They found that higher values of DHI result in a lower occurrence of earlier fibrosis. It remains to be settled whether the LKB NTCP model fairly represents all the available information on breast fibrosis formation. Radiationinduced fibrosis is dependent on several factors, such as the type of surgery performed, the age of the patient, and nutrition. Any model will require some assumptions be made on the range and distribution of these factors. In this paper, we chose a model that allowed us to easily calculate EUD, NTCP, and TCP. However, despite the increasing power of computer technology, NTCP analysis has not seen much use in treatment planning. Nevertheless, NTCP analysis can impart crucial information about radiotherapy-induced complication rates to clinical practice. In a comparative study involving APBI and conventional treatment, Jothy, et al.17 reported that the EUD for APBI was 30.9 Gy. Our EUD is similar and slightly lower than that provided by Bovi et al.10 In the present clinical case, the NTCP of the left breast is higher than that of the right breast, and inversely, the TCP of the left breast is lower. The difference may be caused by cavity size and the type of device (8 –1 in the right breast vs. 6 –1 in the left). Our calculated TCPs are comparable with those reported by Jothy et al.17 Both breasts have very high TCP, which could translate into a good outcome for the patient. The present analysis is limited to breast fibrosis because APBI is believed to reduce breast fibrosis more than conventional wholebreast radiotherapy.2 Still, there is a paucity of clinical data on postradiotherapy breast fibrosis, so validation of the NTCP results is difficult.
Table 1 DVH and radiobiological parameters in the left and right breasts Breasts Parameters 3
PTV (cm ) PTV_EVAL (cm3) CAVITY (cm3) V200 (cm3) V150 (cm3) V100 V95 V90 DHI ODI DNR EUD (Gy) NTCP % TCP %
219
Left
Right
68.49 51.76 16.78 11.77 22.78 47.09 95.75% 100.95% 0.469 0.177 0.333 33.53 4.2 99.2
100.61 70.20 30.40 15.10 29.95 65.96 98.76% 103.52% 0.512 0.150 0.297 29.10 2.57 99.9
Conclusion The SAVI device is gaining acceptance in the management of earlystage breast cancer as a means of improving dose coverage and delivery in limited-duration treatments. The SAVI system accommodates a large pool of patients for whom the MammoSite and Contura devices are unsuitable because of a short distance between the skin or chest wall and the lumpectomy cavity. This study has examined the radiobiological factors of a complex APBI treatment and enhances our understanding of breast fibrosis determinants.
References 1. Fewer, E.J.; Won, L.M.; Boring, C.C.; et al. The lifetime risk of developing breast cancer. J. Natl. Cancer. Inst. 85:892–7; 1993. 2. Vicini, F.; Beitsch, P.D., Quiet, C.A.; et al.Three-year analysis of treatment efficacy, cosmeses, and toxicity by the American Society of Breast Surgeons MammoSite Breast Brachytherapy Registry Trial in patients treated with accelerated partial breast irradiation (APBI). Cancer 112:758 – 66; 2008. 3. Gurdalli, S.; Kuske, R.R.; Quiet, C.A.; et al. Dosimetric performance of Strut-Adjusted volume implant: A new single-entry multicatheter breast brachytherapy applicator. Brachytherapy 10:128 –35; 2011. 4. Meertens, H.; Borger, J.; Steggerda, M.; et al. Evaluation and optimization of interstitial brachytherapy dose distribution. In: Mould RF, Battermann JJ, Martinez AA, et al, editors. Brachytherapy from Radium to Optimization. Veenendaal, The Netherlands: Nucletron International; 1994. pp. 300 – 6. 5. Saw, C.B.; Suntharalingam, N.; Wu, A. Concept of dose nonuniformity in interstitial brachytherapy. Int. J. Radiat. Oncol. Biol. Phys. 26:519 –27; 1993. 6. Seppenwoodle, Y.; Lebesque, J.V.; de Jaeger, K.; et al. Comparing different NTCP model that predict the incidence of radiation pneumonitis. Int. J. Radiat. Oncol. Biol. Phys. 55:724 –35; 2003. 7. Lyman, J.T. Complication probability as assessed from dose volume histograms. Radiat. Res. 104:S13–9; 1985. 8. Burman, C.; Kutcher, G.J.; Emami, B.; et al. Fitting of normal tissue tolerance data to an analytic function. Int. J. Radiat. Oncol. Biol. Phys. 21:123–35; 1991. 9. Emami, B.; Lyman, J.; Brown, A.; et al. Tolerance of normal tissue to therapeutic irradiation. Int. J. Radiat. Oncol. Biol. Phys. 21:109 –22; 1991. 10. Bovi, J.; Qi, X.S.; White, J.; et al. Comparison of three accelerated partial breast irradiation techniques: Treatment effectiveness based upon biological models. Radiother. Oncol. 84:226 –32; 2007. 11. Warkentin, B.; Stavreb, P.; Stravevra, N.; et al. A TCP-NTCP estimation module using DVHs and known radiobiological models and parameters sets. J. Appl. Clin. Med. Phys. 5:50 – 63; 2004. 12. Mancharan, S.R.; Rodriguez, R.R.; Bobba, V.S.; et al. Dosimetry evaluation of SAVI based HDR brachytherapy for partial breast irradiation. J. Med. Phys. 35:131– 6; 2010. 13. Kim, Y.; Trombetta, M.G.; Miften, M. Comparison of single and multiple dwell position methods in MammoSite high dose rate (HDR) brachytherapy planning. J. Appl. Clin. Med. Phys. 11:3235; 2010. 14. Dickler, A.; Kirk, M.; Choo, J.; et al. Treatment volume and dose optimization of MammoSite breast brachytherapy applicator. Int. J. Radiat. Oncol. Biol. Phys. 59: 469 –74; 2004.
220
A.M. Gloi et al. / Medical Dosimetry 37 (2012) 214-220
15. Bloom, S.E.; Kirsner, S.; Mason, B.E.; et al. Accelerated partial breast irradiation using the strut adjusted volume implant single-entry hybrid catheter in brachytherapy for breast cancer in the setting of breast augmentation. Brachytherapy. 10:178 – 83; 2011. 16. Wazer, D.E.; Kramer, B.; Schmid, C.; et al. Factors determining outcome in patients treated with interstitial implantation as a radiation boost for breast conservation therapy. Int. J. Radiat. Oncol. Biol. Phys. 39:381–93; 1997.
17. Jothy Basu, K.S.; Bahl, A.; Subramani, V.; et al. Normal tissue complication probability of fibrosis in radiotherapy of breast cancer: Accelerated partial breast irradiation vs conventional external-beam radiotherapy. J. Cancer Res. Ther. 4:126 –30; 2008. 18. Scanderbeg, D.; Barna, P.; Jiang, S. Dosimetric comparison of SAVI, MammoSite, Contura and Clearpath for accelerated partial breast irradiation. Med. Phys. 36: 2536; 2009.
Medical Dosimetry journal homepage: www.meddos.org
A case report on bilateral partial breast irradiation using SAVI Aime M. Gloi, Ph.D.,* Robert Buchanan M.Sc., D.A.B.R.,† Jeff Nuskind, M.S., D.A.B.R.,* Corrie Zuge, B.S., C.M.D.,* and Anndrea Goettler, C.M.D.* *Radiation Oncology, St. Vincent Hospital, Green Bay, WI; and †Southeast Alabama Medical Center, Radiation Oncology Department, Dothan, AL 36301
A R T I C L E
I N F O
Article history: Received 2 May 2011 Accepted 10 August 2011 Keywords: SAVI DVH DHI ODI DNR EUD NTCP TCP
A B S T R A C T To assess dosimetric parameters in a case study where bilateral accelerated partial breast irradiation (APBI) is delivered using a strut-adjusted volume implant (SAVI) device. A 59-year-old female received APBI in both breasts over 5 days, with fractions of 3.4 Gy twice daily. A Vac-lok system was used for immobilization, and a C-arm was used for daily imaging. We generated dose-volume histograms (DVHs) for the brachytherapy plans to derive several important biologic factors. We calculated the normal tissue complication probability (NTCP), equivalent uniform dose (EUD), and tumor control probability (TCP) using the Lyman-Kutcher-Burman model parameters ␣ ⫽ 0.3 Gy⫺1, ␣/ ⫽ 4 Gy, n ⫽ 0.1, and m ⫽ 0.3. In addition, we assessed the dose homogeneity index (DHI), overdose index, and dose nonuniformity ratio. D95 was ⬎95% and V150 was ⬍50 mL for both breasts. The DHIs were 0.469 and 0.512 for the left and right breasts, respectively. The EUDs (normalized to 3.4 Gy b.i.d.) were 33.53 and 29.10 Gy. The TCPs were estimated at 99.2% and 99.9%, whereas the NTCP values were 4.2% and 2.57%. In this clinical case, we were able to quantify the dosimetric parameters of an APBI treatment performed with a SAVI device. 䉷 2012 American Association of Medical Dosimetrists.
Introduction Breast cancer is one of the most common types of cancer among women in the United States. According to a study by Fewer et al., on average 192,370 women are diagnosed with the disease every year, and approximately 20% of those women will die of it.1 Early-stage breast cancer is characterized by a lack of lymph nodes, metastasis, and clinical lesions with diameters up to 2 cm. In recent years, high public awareness and early mammography screening have increased the number of women diagnosed with early-stage breast cancer. The treatment method of choice for many of these women (40%) is a lumpectomy followed by accelerated partial breast irradiation (APBI).1 This treatment has 2 major advantages compared with other available methods of removing the cancerous tissue: it requires a shorter time commitment (5 days vs several weeks) and it is less invasive. An abundance of devices have been developed for APBI treatment. For example, the Mammosite RTS (Cytyc, Corp., Marlborough, MA) consists of a balloon with a single central lumen. The geometry of the balloon permits the delivery of an increased dose because the radiation is limited to tissue within 1 cm of the balloon’s surface. Good to excellent cosmetic results have been achieved with a MammoSite
catheter when the balloon-to-skin spacing is ⬎7 mm. Skin necrosis and poor cosmesis are observed in several cases when the spacing is ⬍7 mm2. Not all patients are eligible for the device; MammoSite presents some stringent criteria, such as tissue conformance to the balloon, breast size, and other geometrical factors. Following the MammoSite example, the multilumen balloon Contura (SenoRx, Inc., Aliso Viejo, CA) was developed to minimize the radiation “hot spot” in the skin and reduce air/seroma volume. Contura allows for more flexible dosimetry and adapts more easily to geometrical constraints, such as the skin’s distance from the catheter. A hybrid device named SAVI (Strut-Adjusted Volume Implant) from (Cianna Medical, Aliso Viejo, CA), was created to eliminate or minimize the deficiencies of MammoSite and Contura. SAVI provides even more customized radiation therapy, improving the dosimetry and considerably reducing the dose to the skin and chest wall. SAVI also provides more options and precision in its planning system, allowing physicians to tailor the radiation to the tumor site.3 The aim of this study was to analyze the dosimetric parameters of an unusual clinical case, in which the patient received bilateral APBI treatment via a SAVI device. Methods and Materials
Reprint requests to: Aime M. Gloi, Ph.D., Radiation Oncology, St. Vincent Hospital, 835 S. Van Buren Street, PO Box 13508, Green Bay, WI 54307-3508. E-mail: [email protected]
The subject of this report is a 59-year-old female with a breast abnormality noted on a mammogram screening. At the 6-month follow-up visit, the mammogram showed an increase in size and the magnetic resonance image revealed an enhanced area in the
0958-3947/$ – see front matter Copyright 䉷 2012 American Association of Medical Dosimetrists doi:10.1016/j.meddos.2011.08.001
A.M. Gloi et al. / Medical Dosimetry 37 (2012) 214-220
breast. A core biopsy was performed, showing poorly differentiated invasive ductal carcinoma with the globular features typically associated with ductal carcinoma in situ (DCIS). The disease was staged as T2, N0, and M0 grade 3 infiltrating ductal carcinoma of both breasts. The patient was implanted with a SAVI 6 –1 (6 peripheral catheters and one central catheter) device in the left breast, and a SAVI 8 –1 (8 peripheral and one central) in the right breast. The patient was immobilized in a Vac-lok (Medtech, Orange City, IA) cushion marked for reproducible positioning. These cushions are designed for comfortable, patient-specific repositioning; they form a custom mold of the patient’s anatomical contours in any desired pose. The patient’s position during the computed tomography (CT) scan can therefore be duplicated exactly during treatment.
215
The lumpectomy CT images were acquired with the patient immobilized in the treatment position. The slice thickness of the scan was 2.0 mm. The lumpectomy cavity was identified with the help of surgical clips. A fluoroscopy from a C-arm (General Electric OES 9800, Milwaukee, WI) was performed on the day of the CT scan to gather a baseline measurement. The Brachyvision (Varian Medical, Inc., Palo Alto, CA) treatment planning system was used for contouring and planning. Both the cavity and normal tissues, such as skin, ipsilateral lungs, and ribs, were contoured. The planning treatment volume (PTV) was determined by adding a layer 1.0 cm thick around the cavity. In the treatment planning system, we define the volume PTV_EVAL as the PTV minus the cavity. There were 2
Fig. 1. (a, b) SAVI 6-1 inserted in the left breast.
216
A.M. Gloi et al. / Medical Dosimetry 37 (2012) 214-220
Fig. 2. (a, b) SAVI 8-1 inserted in the right breast. treatment planning goals: to limit the skin dose ⬍100% of the prescribed dose and to limit the volume of air/fluid in the cavity during treatment to ⬍10.0% of the PTV. The fractions were separated by an interval of at least 6 hours for radiobiologic purposes. Each breast received 3.4 Gy per fraction, twice daily, for a total of 10 fractions per breast over 5 days. This report uses 3 quality indexes: the dose homogeneity index (DHI), the overdose volume index (ODI), and the dose nonuniformity ratio (DNR). DHI is defined as the volumes that receive a dose of at least 1.0 times the prescription dose (V100) and at least 1.5 times the reference dose (V150).4 Each volume is expressed as a fraction of the target volume PTV and DHI is the difference between them:
DHI ⫽ 共V100 ⫺ V150兲 ⁄ PTV ODI is the ratio of the volume receiving at least twice the prescribed dose to the volume receiving at least the prescribed dose4:
ODI ⫽ V200 ⁄ PTV Finally, we compute DNR as the ratio between the volume that receives at least 1.5 times the prescribed dose to the volume that receives at least the prescribed dose5:
DNR ⫽ V150 ⁄ PTV
A.M. Gloi et al. / Medical Dosimetry 37 (2012) 214-220
217
Fig. 3. Bilateral accelerated partial breast irradiation through SAVI.
We test the normal tissue complication probability (NTCP) model on the occurrence of breast fibrosis. The Lyman-Kutcher-Burman (LKB) model was used to calculate the NTCP. Based on the differential dose-volume histograms (DVHs), we calculate the equivalent uniform dose (EUD)6:
冉兺
EUD ⫽
i
vi vT
冊
共Di兲1⁄n
n
,
where vi is the fractional volume irradiated to dose Di, vT is the total volume of the structure, and n is a tissue-specific model fitting parameter. In this calculation, n is set to 0.1. Given the EUD, we evaluated the NTCP7 for a breast fibrosis as follows:
NTCP ⫽
1
兹2
兰
1
⫺⬁
2⁄2
e⫺x dx
where
t⫽
EUD ⫺ D50 m ⫻ D50
.
D50 is the prescribed dose with a 50% chance of complications, and m is another model parameter inversely proportional to the steepness of the dose response. Here, we use D50 ⫽ 70 Gy and m ⫽ 0.3.8, 9 The tumor control probability (TCP) relies on the fact that all cancer cells have been destroyed.8 It is given by
TCP ⫽ exp 共⫺K ⫻ S兲 , where k ⫽ 200 and the surviving fraction S is given by10
冋冉
S ⫽ exp ⫺ ␣ ⫻ EUD ⫹ d ⫻ EUD ⫺ 0.5␥
EUD d
冊册
.
To calculate S, we take the parameters of an external beam treatment with 2 Gy per fraction: ␣ ⫽ 0.3 Gy⫺1, ␣/ ⫽ 4 Gy, and ␥ ⫽ 0.693/Td (where Td is the tumor cell doubling time, or 15 d).11
Results The patient received a SAVI 6 –1 (Fig. 1a, 1b) in her left breast, and a SAVI 8 –1(Fig. 2a, 2b) in her right breast (Fig. 3). The mean cavity
volumes were 16.8 cm3 for the left breast and 30.4 cm3 for the right breast. The PTV_EVAL volume was greater in the right breast (70.2 cm3) than in the left breast (51.8 cm3). The DVHs are shown in Fig. 4a and 4b. V95, defined as the percentage of PTV_EVAL receiving at least 95% of the prescribed dose, was 95.75% and 98.76% for the left and right breasts, respectively. The skin dose in the left breast reached the limiting value of 100%. The skin dose in the right breast, however, was well below 100%. The DHI was 0.512 for the right breast and 0.469 for the left breast. The ODI was estimated at 0.177 for the left breast and 0.150 for the right breast. NTCP distributions were calculated for both breasts to assess the probability of breast fibrosis, using the values D50 ⫽ 70 Gy and m ⫽ 8, 9 0.3. The estimated probabilities were 4.27% for the left breast and 2.57% for the right breast. Notice that the left breast was implanted with a 6 –1 device, whereas an 8 –1 device was used for the right breast. The TCPs were 99.2% and 99.9% for the left and right breasts, respectively. Discussion All the dosimetric criteria for this APBI treatment were met for both breasts: V95 ⬎95%, V150 ⬍50 mL, and V200 ⬍20 mL. The data for this case compare well with the study of Gurdalli et al.,3 wherein 15 patients were treated with a variety of SAVI devices. Similar dose coverage was observed by Mancharan et al.12 The air/seroma volume was not a factor in this treatment, as is sometimes the case in MammoSite, having been contained by the struts. A SAVI (Fig. 5) device with several struts allows the doctor to shape the dose, resulting in good coverage. It is believed that all 3 devices (MammoSite, Contura, and SAVI) provide similar coverage of the target volume, but SAVI offers the most dose flexibility for patients whose skin and chest wall are close to the cavity.3
218
A.M. Gloi et al. / Medical Dosimetry 37 (2012) 214-220
Fig. 4. (a, b) DHH of right and left breast from top to bottom, respectively, describing the dose coverage of PTV and PTV_Eval.
A quantitative comparison of dose distributions in the 2 breasts is provided in Table 1. For an ideal implant, the quality indexes would be DHI ⫽ 1, ODI ⫽ 0, and DNR ⫽ 0. As anticipated, SAVI achieves a lower DHI than MammoSite: 0.59 (Gurdalli et al.3) in contrast to 0.66 (Yongbok et al.13) or 0.59 in contrast to 0.62 (Dickler et al.14). Because of the greater inhomogeneity of its radiation dose, MammoSite also treated a larger volume of breast tissue (between 110% and 170% of the prescribed dose) than did SAVI. In a recent paper, Gurdalli et al.3 reported results obtained with several SAVI devices. The mean DHI was 0.559 (range 0.52– 0.60),
which is somewhat higher than the indexes shown in Table 1 (0.469 and 0.512 for the left and right breasts, respectively). By contrast, Bloom et al.15 obtained a much lower mean DHI of 0.32. This inconsistent result shows that the methodology of APBI treatments requires more attention. Similar inconsistencies are observed in the DNR and ODI values. In our case, the low values of DNR and ODI suggest that the unnecessary irradiation of normal tissue around the target volume was negligible. Interestingly, Wazer et al.15 uncovered a correlation between DHI and cosmetic outcome in treatments using interstitial LDR boost breast implants.
A.M. Gloi et al. / Medical Dosimetry 37 (2012) 214-220
The available data are both sparse and noisy; consequently, any model predictions based on clinical data will have wide confidence ranges and could be considerably biased. Another problem with the evaluation of TCP and NTCP is the complexity of the radiobiological model and the difficulty of finding adequate fitting parameters. The LKB model certainly does not adequately reflect known radiobiological theory, because the radiation response can be highly dependent on the treatment scheme. In addition, the model assumes a homogeneous organ so it does take into account tissue damage at contiguous sites. Also, it is well understood that the breast DVH may not be consistent during CT and treatment because of variations in pose and the patient’s breathing motion. This problem may cause the overall accuracy to suffer, even if a C-arm is used for daily imaging. In this report, the TCP for posttreatment fibrosis was calculated through the EUD for a nonuniform dose. Our predictions are comparable with those of several authors.3,5,10,16-18 The NTCP and TCP can be calculated using treatment-only variables and the radiobiological characteristics of tissue. Despite the limitations of the model, this case study has provided useful insight into the radiobiological parameters of normal tissue responses.
Fig. 5. SAVI device.
They found that higher values of DHI result in a lower occurrence of earlier fibrosis. It remains to be settled whether the LKB NTCP model fairly represents all the available information on breast fibrosis formation. Radiationinduced fibrosis is dependent on several factors, such as the type of surgery performed, the age of the patient, and nutrition. Any model will require some assumptions be made on the range and distribution of these factors. In this paper, we chose a model that allowed us to easily calculate EUD, NTCP, and TCP. However, despite the increasing power of computer technology, NTCP analysis has not seen much use in treatment planning. Nevertheless, NTCP analysis can impart crucial information about radiotherapy-induced complication rates to clinical practice. In a comparative study involving APBI and conventional treatment, Jothy, et al.17 reported that the EUD for APBI was 30.9 Gy. Our EUD is similar and slightly lower than that provided by Bovi et al.10 In the present clinical case, the NTCP of the left breast is higher than that of the right breast, and inversely, the TCP of the left breast is lower. The difference may be caused by cavity size and the type of device (8 –1 in the right breast vs. 6 –1 in the left). Our calculated TCPs are comparable with those reported by Jothy et al.17 Both breasts have very high TCP, which could translate into a good outcome for the patient. The present analysis is limited to breast fibrosis because APBI is believed to reduce breast fibrosis more than conventional wholebreast radiotherapy.2 Still, there is a paucity of clinical data on postradiotherapy breast fibrosis, so validation of the NTCP results is difficult.
Table 1 DVH and radiobiological parameters in the left and right breasts Breasts Parameters 3
PTV (cm ) PTV_EVAL (cm3) CAVITY (cm3) V200 (cm3) V150 (cm3) V100 V95 V90 DHI ODI DNR EUD (Gy) NTCP % TCP %
219
Left
Right
68.49 51.76 16.78 11.77 22.78 47.09 95.75% 100.95% 0.469 0.177 0.333 33.53 4.2 99.2
100.61 70.20 30.40 15.10 29.95 65.96 98.76% 103.52% 0.512 0.150 0.297 29.10 2.57 99.9
Conclusion The SAVI device is gaining acceptance in the management of earlystage breast cancer as a means of improving dose coverage and delivery in limited-duration treatments. The SAVI system accommodates a large pool of patients for whom the MammoSite and Contura devices are unsuitable because of a short distance between the skin or chest wall and the lumpectomy cavity. This study has examined the radiobiological factors of a complex APBI treatment and enhances our understanding of breast fibrosis determinants.
References 1. Fewer, E.J.; Won, L.M.; Boring, C.C.; et al. The lifetime risk of developing breast cancer. J. Natl. Cancer. Inst. 85:892–7; 1993. 2. Vicini, F.; Beitsch, P.D., Quiet, C.A.; et al.Three-year analysis of treatment efficacy, cosmeses, and toxicity by the American Society of Breast Surgeons MammoSite Breast Brachytherapy Registry Trial in patients treated with accelerated partial breast irradiation (APBI). Cancer 112:758 – 66; 2008. 3. Gurdalli, S.; Kuske, R.R.; Quiet, C.A.; et al. Dosimetric performance of Strut-Adjusted volume implant: A new single-entry multicatheter breast brachytherapy applicator. Brachytherapy 10:128 –35; 2011. 4. Meertens, H.; Borger, J.; Steggerda, M.; et al. Evaluation and optimization of interstitial brachytherapy dose distribution. In: Mould RF, Battermann JJ, Martinez AA, et al, editors. Brachytherapy from Radium to Optimization. Veenendaal, The Netherlands: Nucletron International; 1994. pp. 300 – 6. 5. Saw, C.B.; Suntharalingam, N.; Wu, A. Concept of dose nonuniformity in interstitial brachytherapy. Int. J. Radiat. Oncol. Biol. Phys. 26:519 –27; 1993. 6. Seppenwoodle, Y.; Lebesque, J.V.; de Jaeger, K.; et al. Comparing different NTCP model that predict the incidence of radiation pneumonitis. Int. J. Radiat. Oncol. Biol. Phys. 55:724 –35; 2003. 7. Lyman, J.T. Complication probability as assessed from dose volume histograms. Radiat. Res. 104:S13–9; 1985. 8. Burman, C.; Kutcher, G.J.; Emami, B.; et al. Fitting of normal tissue tolerance data to an analytic function. Int. J. Radiat. Oncol. Biol. Phys. 21:123–35; 1991. 9. Emami, B.; Lyman, J.; Brown, A.; et al. Tolerance of normal tissue to therapeutic irradiation. Int. J. Radiat. Oncol. Biol. Phys. 21:109 –22; 1991. 10. Bovi, J.; Qi, X.S.; White, J.; et al. Comparison of three accelerated partial breast irradiation techniques: Treatment effectiveness based upon biological models. Radiother. Oncol. 84:226 –32; 2007. 11. Warkentin, B.; Stavreb, P.; Stravevra, N.; et al. A TCP-NTCP estimation module using DVHs and known radiobiological models and parameters sets. J. Appl. Clin. Med. Phys. 5:50 – 63; 2004. 12. Mancharan, S.R.; Rodriguez, R.R.; Bobba, V.S.; et al. Dosimetry evaluation of SAVI based HDR brachytherapy for partial breast irradiation. J. Med. Phys. 35:131– 6; 2010. 13. Kim, Y.; Trombetta, M.G.; Miften, M. Comparison of single and multiple dwell position methods in MammoSite high dose rate (HDR) brachytherapy planning. J. Appl. Clin. Med. Phys. 11:3235; 2010. 14. Dickler, A.; Kirk, M.; Choo, J.; et al. Treatment volume and dose optimization of MammoSite breast brachytherapy applicator. Int. J. Radiat. Oncol. Biol. Phys. 59: 469 –74; 2004.
220
A.M. Gloi et al. / Medical Dosimetry 37 (2012) 214-220
15. Bloom, S.E.; Kirsner, S.; Mason, B.E.; et al. Accelerated partial breast irradiation using the strut adjusted volume implant single-entry hybrid catheter in brachytherapy for breast cancer in the setting of breast augmentation. Brachytherapy. 10:178 – 83; 2011. 16. Wazer, D.E.; Kramer, B.; Schmid, C.; et al. Factors determining outcome in patients treated with interstitial implantation as a radiation boost for breast conservation therapy. Int. J. Radiat. Oncol. Biol. Phys. 39:381–93; 1997.
17. Jothy Basu, K.S.; Bahl, A.; Subramani, V.; et al. Normal tissue complication probability of fibrosis in radiotherapy of breast cancer: Accelerated partial breast irradiation vs conventional external-beam radiotherapy. J. Cancer Res. Ther. 4:126 –30; 2008. 18. Scanderbeg, D.; Barna, P.; Jiang, S. Dosimetric comparison of SAVI, MammoSite, Contura and Clearpath for accelerated partial breast irradiation. Med. Phys. 36: 2536; 2009.