- Email: [email protected]
Impact of dosimetric parameters on local control for patients treated with three-dimensional pulsed dose-rate brachytherapy for cervical cancer
Brachytherapy 13 (2014) 326e331
Impact of dosimetric parameters on local control for patients treated with three-dimensional pulsed dose-rate brachytherapy for cervical cancer Sabrina Boyrie1,*, Claire Charra-Brunaud2, Valentin Harter3, Anne Ducassou1, Youlia Kirova4, Isabelle Barillot5, Claude Krzisch6, Philippe Lang7, Marie-Helene Baron8, Xavier Montbarbon9, Martine Delannes1, Didier Peiffert2 1 Department of Radiotherapy, Institut Claudius Regaud, Toulouse, France Department of Radiotherapy, Centre Alexis-Vautrin, Vandoeuvre-les-Nancy, France 3 Department of Biostatistics, Centre Alexis-Vautrin, Vandoeuvre-les-Nancy, France 4 Department of Radiotherapy, Institut Curie, Paris, France 5 Department of Radiotherapy, Centre GF Leclerc, Dijon, France 6 Department of Radiotherapy, Hopital Sud, Amiens, France 7 Department of Radiotherapy, Hopital de la Pitie Salpetriere, Paris, France 8 Department of Radiotherapy, Hopital J Minjoz, Besanc¸on, France 9 Department of Radiotherapy, Centre Leon Berard, Lyon, France
2
ABSTRACT
PURPOSE: To investigate the impact of dose-volume histograms parameters on local control of three-dimensional (3D) image-based pulsed dose-rate brachytherapy (BT). METHODS AND MATERIALS: Within a French multicentric prospective study, the data of the 110 patients treated for cervical cancer with external beam radiotherapy followed by 3D imagebased and optimized pulsed dose-rate BT were analyzed. Delineation procedures were performed on magnetic resonance imaging in a minority of cases and on CT for the majority of cases, adapted from the Gynaecological Groupe Europeen de CurietherapiedEuropean Society for Therapeutic Radiology and Oncology recommendations. Optimization procedure was left to the discretion of the treating center. RESULTS: At 2 years, local control rate reached 78%. Dose to Point A, total reference air kerma, and intermediate-risk clinical target volume (IR-CTV) V60 were predictive factors for local control ( p 5 0.001, p 5 0.001, and p 5 0.013, respectively). Patients with IR-CTV V60 !75% had a relative risk of local recurrence of 3.8 (95% confidence interval, 1.4e11.1). There was no correlation found between the high-risk clinical target volume dosimetric parameters and local control. CONCLUSIONS: This multicentric study has shown that 3D image-based BT provides a high local control rate for cervical cancer patients. The V60 for IR-CTV was identified as an important predictive factor for local control. Ó 2014 American Brachytherapy Society. Published by Elsevier Inc. All rights reserved.
Keywords:
Cervical cancer; Three-dimensional brachytherapy; Pulsed dose-rate brachytherapy; Local control
Introduction Brachytherapy (BT) has played an essential role in the treatment of cervical cancers for decades. Its rapid dose fall off, allowing for very high dose to the central pelvis while Received 7 November 2013; received in revised form 7 February 2014; accepted 7 March 2014. The authors declare that they have no conflict of interest. * Corresponding author. Department of Radiotherapy, Institut Claudius Regaud, 20-24 rue du pont Saint Pierre, 31052 Toulouse cedex, France. Tel.: þ33 561424188; fax: þ33 561424643. E-mail address: [email protected] (S. Boyrie).
relatively sparing bladder, rectum, sigmoid, and small bowel, is key for local control. Achieving local control is a major aim in the care of cervical cancer. Indeed, recurrences occur mainly in the pelvis, and their prognosis is poor. A recent publication using Surveillance Epidemiology and End Result data has shown a 15% decrease in overall survival for cervix carcinoma in the United States parallel to the decline in utilization of BT since 2003 (1). Pulsed dose-rate (PDR) BT was developed toward the end of the 90s and combines the biologic advantages of lowedose-rate BT with the possibility of dose optimization using dwell time position variations. The concept of dose
1538-4721/$ - see front matter Ó 2014 American Brachytherapy Society. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.brachy.2014.03.003
S. Boyrie et al. / Brachytherapy 13 (2014) 326e331
optimization makes sense only if three-dimensional (3D) imaging is integrated into the process of treatment planning. In 2005 and 2006, the Groupe Europeen de Curietherapied European Society for Therapeutic Radiology and Oncology (GEC-ESTRO) provided recommendations for target delineation using 3D imaging (2) and for 3D dose-volume parameters (3). A national nonrandomized multicentric prospective study called ‘‘STIC PDR’’ (Programme de Soutien aux Techniques Innovantes Co^ uteuses Pulsed Dose Rate) was organized for patients treated with curative intent for cervix carcinoma. The study compared a PDR BT method based on 3D imaging with a lowedose-rate BT method based on orthogonal two-dimensional (2D) X-rays. The results showed an improvement of local control with lesser Grade 3e4 toxicities in the 3D arm (4, 5). The aim of the present study was to investigate if dose-volume histograms (DVH) in patients treated with 3D BT can predict local control.
327
Table 1 Patient and tumor characteristics Number of patients Age (y), mean SD Median age (minemax) Histology, n (%) Squamous cell Adenocarcinoma Other FIGO stage, n (%) IB1 IB2 IIA IIB IIIA IIIB Tumor size (mm), mean SD Pelvic nodes,a n (%) Lombo-aortic nodesa
110 53.3 14.1 51 (25e84) 93 (84) 6 (14) 1 (1) 11 (10) 16 (15) 11 (10) 45 (41) 5 (5) 21 (19) 48.6 15.8 48 (44) 13 (12)
FIGO 5 Federation of Gynecology and Obstetrics; min 5 minimum; max 5 maximum. a Nodes diagnosed on imagery (CT or MRI).
Methods and materials Patient and tumor characteristics Between January 2005 and June 2007, 801 patients were enrolled into STIC PDR through 20 participating centers. For the present report, we analyzed the 110 patients of this trial who received external beam radiation therapy (EBRT) chemotherapy followed by a single application of BT. For all patients, the histologic diagnosis of cervical cancer was obtained from pretreatment biopsies. The initial locoregional staging examination included pelvic MRI for all patients, associated with thoraco-abdomino-pelvic CT scan or positron emission tomography CT scan. The disease was staged according to the 1994 International Federation of Gynecology and Obstetrics (FIGO) classification. The patients were informed about their inclusion in the study. Patients had squamous cell carcinoma (84%) or adenocarcinoma (14%) with FIGO stages IB1 to IVB (10% Stage IB1, 65% stages IB2/IIA/IIB, and 24% stages IIIA/IIIB); 48 patients had pelvic nodal involvement and 13 lumbo-aortic nodal involvement. Patients’ characteristics are summarized in Table 1. Treatment characteristics EBRT consisted of 3D conformal radiotherapy using a four-field technique delivering 45 Gy in 25 fractions and 5 weeks to the pelvis. Para-aortic irradiation was applied in case of nodal involvement. A nodal or parametrial boost was allowed in case of involvement. Concomitant chemotherapy based on cisplatin 5FU was proposed to all patients with stage $IB2, in the absence of contraindication. One or 2 weeks after completion of EBRT, patients were scheduled for BT. It was recommended that the overall
treatment time, from the beginning of EBRT to the end of BT, to be kept under 8 weeks (6). At the beginning of the BT procedure, a careful clinical examination was performed to assess the tumor extension and topography. A Foley urinary catheter was inserted and fixed against the bladder neck after the bladder balloon was filled with 7 mL of diluted iodine. The choice of the applicator was left to the discretion of the treating center (Fletcher applicator, vaginal mold, tandem and ring, etc.). Treatment planning was performed based on CT or MRI scans. After implantation, the patients were transferred to the CT or MRI scanner, and compatible dummy sources were inserted into the catheters to visualize the intrauterine and vaginal sources; 3-mm slice thickness was recommended; bladder filling with iodine was optional as was intravenous opacification. Type and number of sequences for MRI were performed according to the local procedures. Axial postinsertion pelvic images were imported into the treatment planning system, and a 3D set was reconstructed. Delineation procedures were adapted from GEC-ESTRO recommendations (2). Because of poor visualization, the GTV was not delineated on CT scans. High-risk clinical target volume (HR-CTV) included the whole cervix plus macroscopic disease at the time of BT. Intermediate-risk clinical target volume (IR-CTV) was delineated encompassing HR-CTV plus a variable margin depending on the initial extent of the disease and response to initial treatment. For IB1 and IB2 tumors, the IR-CTV comprised a margin of 1 cm toward the vagina, the uterine corpus, and the parametrium. For more advanced stages, the IRCTV was delineated at least up to the initial extent of the disease in case of good response to EBRT, or further in case of poor response, to have a 1-cm margin around the residual disease at the time of BT. To identify the dwell positions, the dummy source images were digitized on each axial
328
S. Boyrie et al. / Brachytherapy 13 (2014) 326e331
CT or MRI slice. The length and position of the sources were adapted to the patient’s anatomy and to the tumor extent. All patients were treated with PDR BT. Active lengths of the vaginal and uterine sources were determined, and the dose was generally prescribed to Point A as a starting point. Manual or graphical optimization was then performed to lower the dose to the organs at risk (OARs) and/or improve the dose to the CTVs. The aim was to keep the ratios of consecutive active dwell time positions below 2. The following DVH parameters were calculated for HR and IR-CTV: dose received by 100% and 90% of the volume (D100 and D90) and percentage of the volume receiving 60 and 85 Gy (V60 and V85), respectively. Dose to Point A and total reference air kerma were also reported. Physical doses delivered to the target volumes were converted into biologically effective doses for more accurate comparisons and normalized to a radiobiologically weighted dose equivalent of 2 Gy/fraction (EQD2) (3, 7, 8). The linearequadratic model was applied using a/b of 10 Gy for target volumes and sublethal damage repair half-time of 1.5 hours for both (9e11). Normalized doses are denoted by Gya/b10 for the target doses. A total dose (EBRT þ BT) of 60 Gy was recommended. This dose prescription corresponded to the volume encompassed by the 60-Gy reference volume as defined by the International Commission on Radiation Units 38 report. For patients receiving 45 Gy of EBRT to the pelvis, a BT boost prescription of 15 Gy was recommended for the IR-CTV D90. The dose to the HR-CTV D90 aimed to reach 85 Gy but was limited by the dose-volume constraints for the OARs according to local protocols. There were no protocol recommendations for dose constraints to the OARs, but recommendations were established for dose rate (12). If the dose rate to one of the OAR D2cm3 was more than 0.6 Gy/h, the dose rate at the prescribed isodose was decreased and the total number of pulses was increased to reach the same total dose. After validation of the optimized plan, PDR BTwas initiated, delivering one pulse per hour, 24 h/d. Followup of patients comprised a gynecologic examination 2 months after BT and then every 6 months for 2 years. Systematic CT or MRI was recommended at 12 and 24 months after treatment completion. Statistical analyses A description of demographic, histologic, and clinical data was performed and presented as mean (SD) for quantitative variables, as frequency tables for qualitative ones. Local recurrence-free survival was defined as relapse occurring in the CTV and locoregional recurrence-free survival as a relapse occurring in the CTVor nodal pelvic recurrence; overall survival was defined as a death from any cause. Local recurrence-free survival was analyzed using the Cox proportional hazard regression model. The effect of DVH parameters on local control was measured in multivariate Cox proportional hazard models adjusted for a selection of
demographic, histologic, and clinical prognostic factors. Wald’s tests on Cox proportional hazard models were then performed to highlight the specific effect of each of these parameters. Prognostic factors including age, tumor size, pelvic lymph nodes, and para-aortic lymph nodes were selected by a stepwise procedure modeling local recurrence-free survival on Cox proportional hazard multivariate models. Note that the FIGO stage was not selected by this procedure as it is strongly statistically linked to the combination of tumor size and lymph node involvement. Threshold doses that maximize the probability of local recurrence were determined by maximum likelihood estimation on Cox proportional hazard models. A p-value #0.05 was considered statistically significant. All calculations were performed with R software version 2.11.0 (R Foundation for Statistical Computing, Vienna, Austria). Results Treatment The pelvic EBRT was delivered with a four-field technique. Median pelvic central dose was 45.5 Gy (2.6 Gy) in 25 fractions. EBRT included para-aortic irradiation for 18 patients (16%) and a parametrial boost ($8 Gy) for 19 of them (17%). Concomitant chemotherapy, mainly weekly cisplatin, was delivered to 90% of patients. BT treatment planning was based on CT scan for 87% of the patients and on MRI for 13%. Dosimetric parameters are listed in Table 2. The mean dose to Point A was 70 Gya/b10 (9.6 Gy). The mean HR-CTV D90 was 73.2 Gya/b10 (11.3 Gy). The mean V85 for HR-CTV was 59% (25). The mean IR-CTV D90 was 61.8 Gya/b10 (6.9 Gy), and the mean V60 for IR-CTV was 86% (20). Table 2 Brachytherapy dosimetric data
Dosimetric parameters
Mean SD
Adjusted hazard of local recurrence p-valuea
Dose to Point A (Gya/b10) TRAK (Gy$cm2) HR-CTV Volume (cm3) D90 (Gya/b10) D100 (Gya/b10) V85 (%) IR-CTV Volume (cm3) D90 (Gya/b10) D100 (Gya/b10) V60 (%)
70 9.6 184 105
0.05 0.001
35 73.2 62.0 59
27 11.3 8.0 25
0.95 0.18 0.28 0.23
99 61.8 54.4 86
56 6.9 4.7 20
0.53 0.39 0.09 0.01
TRAK 5 total reference air kerma; HR-CTV 5 high-risk clinical target volume; IR-CTV 5 intermediate-risk clinical target volume; D90 5 dose received by 90% of target volume; D100 5 minimal target dose; V60 5 percentage of target volume receiving 60 Gy; V85 5 percentage of target volume receiving 85 Gy. a Wald test p-value in Cox proportional hazard model adjusted for age, tumor size, pelvic lymph nodes, and para-aortic lymph nodes.
S. Boyrie et al. / Brachytherapy 13 (2014) 326e331
The median overall treatment time (from the EBRT first session to BT) was 55 days (15). Clinical outcome Median followup was 24.2 months (range, 2.5e39.1); 85% of the patients had a followup time of more than 20 months or until they relapsed. At 24 months, local recurrence-free survival reached 78% (71; 87) and locoregional recurrence-free survival 70% (61; 79). The median delay to local relapse was 10 months (range, 3.7e24.4 months). Overall survival at 24 months was 74% (70; 83). Those results are accurately described elsewhere (4). Dosimetric predictive factors for local control Well established for 2D treatment planning, dose to Point A and total reference air kerma were confirmed to be predictive factors for local control in our study ( p 5 0.001 and p 5 0.001). Patients with a dose to Point A !68 Gya/b10 had a relative risk of local recurrence of 6.1 (95% confidence interval, 1.7e21.7). Concerning 3D DVH parameters, IR-CTV V60 was in our study significantly lower for patients with local relapse than for patients with local tumor control ( p 5 0.013). Patients with an IR-CTV V60 !75% had a relative risk of local recurrence of 3.8 (95% confidence interval, 1.4e11.1) (Fig. 1). HR-CTV D90 and D100 had no significant impact on local control, nor did HR-CTV V85. No significant dependence of local control on doses per pulse for both HRCTV and IR-CTV was found.
Fig. 1. Local recurrence-free survival curves according to IR-CTV V60. IR-CTV 5 intermediate-risk clinical target volume; V60 5 percentage of target volume receiving 60 Gy.
329
Dosimetric study of local relapses Eighteen patients developed a local recurrence. Dosimetric parameters of these patients are listed in Table 3. Mean dose to Point A was 67.3 Gya/b10 (8.8 Gy). The mean HR-CTV D90 was 69.8 Gya/b10 (11.4 Gy), and the mean V85 for HR-CTV was 50% (26). The mean IR-CTV D90 was 59.4 Gya/b10 (7.8 Gy), and the mean V60 for IR-CTV was 75% (26). Ten complete files with local recurrence could be recovered from the different centers involved in the study, including imaging performed at the time of local relapse (patients 2, 5, 6, 7, 8, 12, 13, 16, 17, and 18). When comparing BT treatment planning and imaging performed at the time of local relapse, we found that nine local relapses occurred within the HR-CTV volume. Six patients had target volumes with insufficient dose coverage (mean HR-CTV D90 5 64.2 Gy 7.8 Gy). Among them, 3 patients had a parametrial involvement poorly covered because of using intracavitary BT alone without interstitial needles. For the 3 other patients who presented a local relapse within the HR-CTV, no dosimetric explanation could be given (mean HR-CTV D90 5 85.3 Gy 1.9 Gy). Only 1 patient had an isolated vaginal relapse outside IR-CTV. It is noted that for all these patients, doses to OAR were easily respected.
Discussion Currently, the impact of dosimetry optimization of 3D image-based BT on clinical outcome has been investigated in only a few studies. Dosimetric parameters that were previously reported as predictive factors for local control included doses delivered to the HR-CTV (HR-CTV D100 and particularly HR-CTV D90). Our study is, to our knowledge, the first one to show that IR-CTV coverage is important for tumor control. We have shown in this study that IR-CTV V60 was predictive for local control ( p 5 0.013). Dimopoulos et al (13) have studied the doseeeffect relationship for local control of locally advanced cervical cancer treated with MRI-optimized treatment planning in combination with interstitial BT. His team reported a trend toward significance for local control for the IR-CTV D90, whereas HR-CTV D100 and HR-CTV D90 were clearly significant. Tumor control rates higher than 90% at 3 years could be expected at doses O86 Gya/b10 and 67 Gya/b10 for D90 and D100, respectively. Similar results were reported by Beriwal et al (14) for Pittsburgh with a local control rate of 88% at 2 years using a hybrid MRI/CT approach (mean HR-CTV D90 5 84 Gy) and by Nomden et al (15) for Utrecht achieving local control rates of 93% at 3 years using MRI-guided BT (mean HR-CTV D90 5 83.3 Gy). Delivering such doses to HR-CTV D90 is necessary but not always sufficient to achieve local control. Because the D90 does not reflect the dose to the whole target, lowdose regions within the target may not be apparent. Others have shown that the minimum point dose is also important
330
S. Boyrie et al. / Brachytherapy 13 (2014) 326e331
Table 3 Dosimetric data of patients with local recurrence
Patients 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Mean SD Median (minemax)
FIGO Tumor stage size (mm)
Pelvic lymph HR-CTV TRAK nodes volume (cm3) (Gy.cm2)
Dose to Point A (Gya/b10)
HR-CTV D90 HR-CTV (Gya/b10) V85 (%)
IR-CTV D90 IR-CTV (Gya/b10) V60 (%)
IB1 IIIB IIIB IIB IIIB IB2 IIIB IIB IIA IIB IIIB IB2 IIB IIIA IB2 IB2 IIB IIB
þ þ þ þ þ þ þ þ þ þ þ
56 59 63 63 65 64 70 59 67 64 65 68 67 65 92 84 65 76 67 9 65 (56e92)
51 54 56 58 59 63 63 68 70 70 73 75 77 80 82 84 84 88 70 11 70 (51e88)
na 54 53 52 59 60 54 62 63 68 55 62 45 na 65 53 72 73 59 8 59 (45e73)
40 80 60 30 50 80 na 65 31 na na 55 80 75 60 60 50 48 58 17 60 (30e80)
38 41 67 27 48 36 29 10 34 71 21 17 14 37 106 25 14 15 36 24 32 (10e106)
60 124 111 104 162 132 125 69 148 173 143 164 157 222 320 157 174 225 154 60 152 (60e225)
8 19 24 33 29 34 35 29 43 48 55 62 78 93 70 78 84 87 50 26 45 (8e93)
na 74 69 59 74 95 71 95 96 99 83 29 60 na 99 95 10 100 75 26 78 (10e100)
FIGO 5 Federation of Gynecology and Obstetrics; TRAK 5 total reference air kerma; HR-CTV 5 high-risk clinical target volume; IRCTV 5 intermediate-risk clinical target volume; na 5 nonavailable; D90 5 dose received by 90% of target volume; V60 5 percentage of target volume receiving 60 Gy; V85 5 percentage of target volume receiving 85 Gy; min 5 minimum; max 5 maximum.
to consider for local control. In their study, Schmid et al (16) found a mean minimum point dose to the HR-CTV of 72 Gya/b10 for patients with local recurrences as opposed to 99 Gya/b10 in the matched group of patients with complete local remission. In our study, the HR-CTV dosimetric factors considered here were not found to be predictive of local control. The absence of a doseeresponse relationship between HR-CTV dose-volume parameters and local control in our study can be partially explained by the relatively low doses delivered to the HR-CTV. Indeed, the mean HRCTV D90 was 73.2 Gya/b10, which is lower than in previous studies. There are several possible explanations for this observation. The main aim of the study was to lower the complication rates. So most centers tried to remain within ICRU maximal doses as 3D OARs constraints were not well established at the time of the study. For most centers, 3D planning for gynecologic BT was a new technique. During this learning period, participating centers carefully moved from 2D to 3D planning without necessarily aiming at increasing doses at the same time. Each center was free to optimize treatment planning according to previous procedures as the study started recruiting in January 2005 before GEC-ESTRO recommendations were published. Thus, the study design did not include major dosimetric constraints. A flaw of this multicentric study is heterogeneity of practice, with disparate doses to target volumes and large SDs. But it reflects dose-prescribing practices across the whole country at that period of time. Finally, these low doses may also be partly explained by using CT scans
for delineation of target volumes. Delineation was mostly based on CT scans (87%) as most centers did not have the possibility to perform MRI for dosimetry at the time of BT. MRI is recommended as the standard imaging technique to be used for the initial evaluation of local spread and has been found to be superior to CT in parametrial evaluation and in tumor size assessment (17, 18). CT tumor contours have been found to overestimate the tumor width compared with MRI-based contours (19, 20). This can result in artificially lower D90 and D100 with CT plans compared with MRI-guided dosimetry. Conclusions This multicentric study reports high local control rates with 3D-optimized BT. It is the first study to demonstrate the importance of IR-CTV coverage on local control. Indeed, IR-CTV V60 was identified as a predictive factor for local control. HR-CTV DVHs were not predictive for local control, perhaps, because relatively low doses were delivered to that volume as the main aim of the study was to reduce the risk of severe late complications. The objective of a phase II on going trial is now to deliver higher doses to CTVs as it has been shown to improve dramatically local control even in cases of locally advanced disease. Acknowledgments We wish to thank Astrid de Leeuw for the material support for the calculation of the radiobiologically equivalent
S. Boyrie et al. / Brachytherapy 13 (2014) 326e331
dose of PDR BT. We thank Dr Alain Haddad for his contribution in the translation and editing of this article. The following radiation oncologists and physicists are being acknowledged for their contribution to the STIC PDR study: J. Bonnet, P. Magnenet, N. Allieres, T. Lacornerie, Y. Kirova, J. Coulot, P. Romestaing, M.-P. Sotton, P. Lang, G. Boisserie, S. Racadot, C. Malet, A. Chemin, D.Williaume, J. Bellec, C. Krzisch, K. Peignaux, J.-P. Brenier, A. Cussac, A. Lisbona, M.-H. Baron, O. Goubard, T.D. Nguyen, N. Gaillot, F. Lesaunier, V. Talguen, and I. Barillot, B. Dubray. The study was supported by a grant from the French Minister of Health and Sports : ‘‘STIC PDR: Utilisation de la curietherapie pulsee gynecologique (PDR) avec optimisation de la repartition de la dose et dosimetrie tridimensionnelle’’ (Ministere de la sante et des sports, Programme de soutien aux innovations diagnostiques et therapeutiques co^uteuses). References [1] Han K, Milosevic M, Fyles A, et al. Trends in the utilization of brachytherapy in cervical cancer in the United States. Int J Radiat Oncol Biol Phys 2013;87:111e119. [2] Haie-Meder C, Potter R, Van Limbergen E, et al. Recommendations from Gynaecological (GYN) GEC-ESTRO Working Group (I): Concepts and terms in 3D image based 3D treatment planning in cervix cancer brachytherapy with emphasis on MRI assessment of GTV and CTV. Radiother Oncol 2005;74:235e245. [3] Potter R, Haie-Meder C, Van Limbergen E, et al. Recommendations from gynaecological (GYN) GEC ESTRO working group (II): Concepts and terms in 3D image-based treatment planning in cervix cancer brachytherapy-3D dose volume parameters and aspects of 3D image-based anatomy, radiation physics, radiobiology. Radiother Oncol 2006;78:67e77. [4] Charra-Brunaud C, Harter V, Delannes M, et al. Impact of 3D imagebased PDR brachytherapy on outcome of patients treated for cervix carcinoma in France: Results of the French STIC prospective study. Radiother Oncol 2012;103:305e313. [5] Levitchi M, Charra-Brunaud C, Quetin P, et al. Impact of dosimetric and clinical parameters on clinical side effects in cervix cancer patients treated with 3D pulse-dose-rate intracavitary brachytherapy. Radiother Oncol 2012;103:314e321. [6] Girinsky T, Rey A, Roche B, et al. Overall treatment time in advanced cervical carcinomas: A critical parameter in treatment outcome. Int J Radiat Oncol Biol Phys 1993;27:1051e1056. [7] Lang S, Nulens A, Briot E, et al. Intercomparison of treatment concepts for MR image assisted brachytherapy of cervical carcinoma
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18] [19]
[20]
331
based on GYN GEC-ESTRO recommendations. Radiother Oncol 2006;78:185e193. Van de Kamer JB, De Leeuw AA, Moerland MA, et al. Determining DVH parameters for combined external beam and brachytherapy treatment: 3D biological dose adding for patients with cervical cancer. Radiother Oncol 2010;94:248e253. Fowler JF, Van Limbergen EF. Biological effect of pulsed dose rate brachytherapy with stepping sources if short half-times of repair are present in tissues. Int J Radiat Oncol Biol Phys 1997;37: 877e883. Visser AG, van den Aardweg GJ, Levendag PC. Pulsed dose rate and fractionated high dose rate brachytherapy: Choice of brachytherapy schedules to replace low dose rate treatments. Int J Radiat Oncol Biol Phys 1996;34:497e505. Dale RG, Coles IP, Deehan C, et al. Calculation of integrated biological response in brachytherapy. Int J Radiat Oncol Biol Phys 1997;38: 633e642. Lambin P, Gerbaulet A, Kramar A, et al. Phase III trial comparing two low dose rates in brachytherapy of cervix carcinoma: Report at two years. Int J Radiat Oncol Biol Phys 1993;25:405e412. Dimopoulos JC, Potter R, Lang S, et al. Dose-effect relationship for local control of cervical cancer by magnetic resonance image-guided brachytherapy. Radiother Oncol 2009;93:311e315. Beriwal S, Kannan N, Kim H, et al. Three-dimensional high dose rate intracavitary image-guided brachytherapy for the treatment of cervical cancer using a hybrid magnetic resonance imaging/computed tomography approach: Feasibility and early results. Clin Oncol (R Coll Radiol) 2011;23:685e690. Nomden CN, de Leeuw AA, Roesnink JM, et al. Clinical outcome and dosimetric parameters of chemo-radiation including MRI guided adaptative brachytherapy with tandem-ovoid applicators for cervical cancer patients: A single institution experience. Radiother Oncol 2013;107:69e74. Schmid MP, Kirisits C, Nesvacil N, et al. Local recurrences in cervical cancer patients in the setting of image-guided brachytherapy: A comparison of spatial dose distribution within a matched-pair analysis. Radiother Oncol 2011;100:468e472. Kim SH, Choi BI, Han JK, et al. Preoperative staging of uterine cervical carcinoma: Comparison of CT and MRI in 99 patients. J Comput Assist Tomogr 1993;17:633e640. Kim SH, Choi BI, Lee HP, et al. Uterine cervical carcinoma: Comparison of CT and MR findings. Radiology 1990;175:45e51. Viswanathan AN, Dimopoulos J, Kirisits C, et al. Computed tomography versus magnetic resonance imaging-based contouring in cervical cancer brachytherapy: Results of a prospective trial and preliminary guidelines for standardized contours. Int J Radiat Oncol Biol Phys 2007;68:491e498. Hegazy N, P€otter R, Kirisits C, et al. High-risk clinical target volume delineation in CT-guided cervical cancer brachytherapy: Impact of information from FIGO stage with or without systematic inclusion of 3D documentation of clinical gynecological examination. Acta Oncol 2013;53:1345e1352.
Impact of dosimetric parameters on local control for patients treated with three-dimensional pulsed dose-rate brachytherapy for cervical cancer Sabrina Boyrie1,*, Claire Charra-Brunaud2, Valentin Harter3, Anne Ducassou1, Youlia Kirova4, Isabelle Barillot5, Claude Krzisch6, Philippe Lang7, Marie-Helene Baron8, Xavier Montbarbon9, Martine Delannes1, Didier Peiffert2 1 Department of Radiotherapy, Institut Claudius Regaud, Toulouse, France Department of Radiotherapy, Centre Alexis-Vautrin, Vandoeuvre-les-Nancy, France 3 Department of Biostatistics, Centre Alexis-Vautrin, Vandoeuvre-les-Nancy, France 4 Department of Radiotherapy, Institut Curie, Paris, France 5 Department of Radiotherapy, Centre GF Leclerc, Dijon, France 6 Department of Radiotherapy, Hopital Sud, Amiens, France 7 Department of Radiotherapy, Hopital de la Pitie Salpetriere, Paris, France 8 Department of Radiotherapy, Hopital J Minjoz, Besanc¸on, France 9 Department of Radiotherapy, Centre Leon Berard, Lyon, France
2
ABSTRACT
PURPOSE: To investigate the impact of dose-volume histograms parameters on local control of three-dimensional (3D) image-based pulsed dose-rate brachytherapy (BT). METHODS AND MATERIALS: Within a French multicentric prospective study, the data of the 110 patients treated for cervical cancer with external beam radiotherapy followed by 3D imagebased and optimized pulsed dose-rate BT were analyzed. Delineation procedures were performed on magnetic resonance imaging in a minority of cases and on CT for the majority of cases, adapted from the Gynaecological Groupe Europeen de CurietherapiedEuropean Society for Therapeutic Radiology and Oncology recommendations. Optimization procedure was left to the discretion of the treating center. RESULTS: At 2 years, local control rate reached 78%. Dose to Point A, total reference air kerma, and intermediate-risk clinical target volume (IR-CTV) V60 were predictive factors for local control ( p 5 0.001, p 5 0.001, and p 5 0.013, respectively). Patients with IR-CTV V60 !75% had a relative risk of local recurrence of 3.8 (95% confidence interval, 1.4e11.1). There was no correlation found between the high-risk clinical target volume dosimetric parameters and local control. CONCLUSIONS: This multicentric study has shown that 3D image-based BT provides a high local control rate for cervical cancer patients. The V60 for IR-CTV was identified as an important predictive factor for local control. Ó 2014 American Brachytherapy Society. Published by Elsevier Inc. All rights reserved.
Keywords:
Cervical cancer; Three-dimensional brachytherapy; Pulsed dose-rate brachytherapy; Local control
Introduction Brachytherapy (BT) has played an essential role in the treatment of cervical cancers for decades. Its rapid dose fall off, allowing for very high dose to the central pelvis while Received 7 November 2013; received in revised form 7 February 2014; accepted 7 March 2014. The authors declare that they have no conflict of interest. * Corresponding author. Department of Radiotherapy, Institut Claudius Regaud, 20-24 rue du pont Saint Pierre, 31052 Toulouse cedex, France. Tel.: þ33 561424188; fax: þ33 561424643. E-mail address: [email protected] (S. Boyrie).
relatively sparing bladder, rectum, sigmoid, and small bowel, is key for local control. Achieving local control is a major aim in the care of cervical cancer. Indeed, recurrences occur mainly in the pelvis, and their prognosis is poor. A recent publication using Surveillance Epidemiology and End Result data has shown a 15% decrease in overall survival for cervix carcinoma in the United States parallel to the decline in utilization of BT since 2003 (1). Pulsed dose-rate (PDR) BT was developed toward the end of the 90s and combines the biologic advantages of lowedose-rate BT with the possibility of dose optimization using dwell time position variations. The concept of dose
1538-4721/$ - see front matter Ó 2014 American Brachytherapy Society. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.brachy.2014.03.003
S. Boyrie et al. / Brachytherapy 13 (2014) 326e331
optimization makes sense only if three-dimensional (3D) imaging is integrated into the process of treatment planning. In 2005 and 2006, the Groupe Europeen de Curietherapied European Society for Therapeutic Radiology and Oncology (GEC-ESTRO) provided recommendations for target delineation using 3D imaging (2) and for 3D dose-volume parameters (3). A national nonrandomized multicentric prospective study called ‘‘STIC PDR’’ (Programme de Soutien aux Techniques Innovantes Co^ uteuses Pulsed Dose Rate) was organized for patients treated with curative intent for cervix carcinoma. The study compared a PDR BT method based on 3D imaging with a lowedose-rate BT method based on orthogonal two-dimensional (2D) X-rays. The results showed an improvement of local control with lesser Grade 3e4 toxicities in the 3D arm (4, 5). The aim of the present study was to investigate if dose-volume histograms (DVH) in patients treated with 3D BT can predict local control.
327
Table 1 Patient and tumor characteristics Number of patients Age (y), mean SD Median age (minemax) Histology, n (%) Squamous cell Adenocarcinoma Other FIGO stage, n (%) IB1 IB2 IIA IIB IIIA IIIB Tumor size (mm), mean SD Pelvic nodes,a n (%) Lombo-aortic nodesa
110 53.3 14.1 51 (25e84) 93 (84) 6 (14) 1 (1) 11 (10) 16 (15) 11 (10) 45 (41) 5 (5) 21 (19) 48.6 15.8 48 (44) 13 (12)
FIGO 5 Federation of Gynecology and Obstetrics; min 5 minimum; max 5 maximum. a Nodes diagnosed on imagery (CT or MRI).
Methods and materials Patient and tumor characteristics Between January 2005 and June 2007, 801 patients were enrolled into STIC PDR through 20 participating centers. For the present report, we analyzed the 110 patients of this trial who received external beam radiation therapy (EBRT) chemotherapy followed by a single application of BT. For all patients, the histologic diagnosis of cervical cancer was obtained from pretreatment biopsies. The initial locoregional staging examination included pelvic MRI for all patients, associated with thoraco-abdomino-pelvic CT scan or positron emission tomography CT scan. The disease was staged according to the 1994 International Federation of Gynecology and Obstetrics (FIGO) classification. The patients were informed about their inclusion in the study. Patients had squamous cell carcinoma (84%) or adenocarcinoma (14%) with FIGO stages IB1 to IVB (10% Stage IB1, 65% stages IB2/IIA/IIB, and 24% stages IIIA/IIIB); 48 patients had pelvic nodal involvement and 13 lumbo-aortic nodal involvement. Patients’ characteristics are summarized in Table 1. Treatment characteristics EBRT consisted of 3D conformal radiotherapy using a four-field technique delivering 45 Gy in 25 fractions and 5 weeks to the pelvis. Para-aortic irradiation was applied in case of nodal involvement. A nodal or parametrial boost was allowed in case of involvement. Concomitant chemotherapy based on cisplatin 5FU was proposed to all patients with stage $IB2, in the absence of contraindication. One or 2 weeks after completion of EBRT, patients were scheduled for BT. It was recommended that the overall
treatment time, from the beginning of EBRT to the end of BT, to be kept under 8 weeks (6). At the beginning of the BT procedure, a careful clinical examination was performed to assess the tumor extension and topography. A Foley urinary catheter was inserted and fixed against the bladder neck after the bladder balloon was filled with 7 mL of diluted iodine. The choice of the applicator was left to the discretion of the treating center (Fletcher applicator, vaginal mold, tandem and ring, etc.). Treatment planning was performed based on CT or MRI scans. After implantation, the patients were transferred to the CT or MRI scanner, and compatible dummy sources were inserted into the catheters to visualize the intrauterine and vaginal sources; 3-mm slice thickness was recommended; bladder filling with iodine was optional as was intravenous opacification. Type and number of sequences for MRI were performed according to the local procedures. Axial postinsertion pelvic images were imported into the treatment planning system, and a 3D set was reconstructed. Delineation procedures were adapted from GEC-ESTRO recommendations (2). Because of poor visualization, the GTV was not delineated on CT scans. High-risk clinical target volume (HR-CTV) included the whole cervix plus macroscopic disease at the time of BT. Intermediate-risk clinical target volume (IR-CTV) was delineated encompassing HR-CTV plus a variable margin depending on the initial extent of the disease and response to initial treatment. For IB1 and IB2 tumors, the IR-CTV comprised a margin of 1 cm toward the vagina, the uterine corpus, and the parametrium. For more advanced stages, the IRCTV was delineated at least up to the initial extent of the disease in case of good response to EBRT, or further in case of poor response, to have a 1-cm margin around the residual disease at the time of BT. To identify the dwell positions, the dummy source images were digitized on each axial
328
S. Boyrie et al. / Brachytherapy 13 (2014) 326e331
CT or MRI slice. The length and position of the sources were adapted to the patient’s anatomy and to the tumor extent. All patients were treated with PDR BT. Active lengths of the vaginal and uterine sources were determined, and the dose was generally prescribed to Point A as a starting point. Manual or graphical optimization was then performed to lower the dose to the organs at risk (OARs) and/or improve the dose to the CTVs. The aim was to keep the ratios of consecutive active dwell time positions below 2. The following DVH parameters were calculated for HR and IR-CTV: dose received by 100% and 90% of the volume (D100 and D90) and percentage of the volume receiving 60 and 85 Gy (V60 and V85), respectively. Dose to Point A and total reference air kerma were also reported. Physical doses delivered to the target volumes were converted into biologically effective doses for more accurate comparisons and normalized to a radiobiologically weighted dose equivalent of 2 Gy/fraction (EQD2) (3, 7, 8). The linearequadratic model was applied using a/b of 10 Gy for target volumes and sublethal damage repair half-time of 1.5 hours for both (9e11). Normalized doses are denoted by Gya/b10 for the target doses. A total dose (EBRT þ BT) of 60 Gy was recommended. This dose prescription corresponded to the volume encompassed by the 60-Gy reference volume as defined by the International Commission on Radiation Units 38 report. For patients receiving 45 Gy of EBRT to the pelvis, a BT boost prescription of 15 Gy was recommended for the IR-CTV D90. The dose to the HR-CTV D90 aimed to reach 85 Gy but was limited by the dose-volume constraints for the OARs according to local protocols. There were no protocol recommendations for dose constraints to the OARs, but recommendations were established for dose rate (12). If the dose rate to one of the OAR D2cm3 was more than 0.6 Gy/h, the dose rate at the prescribed isodose was decreased and the total number of pulses was increased to reach the same total dose. After validation of the optimized plan, PDR BTwas initiated, delivering one pulse per hour, 24 h/d. Followup of patients comprised a gynecologic examination 2 months after BT and then every 6 months for 2 years. Systematic CT or MRI was recommended at 12 and 24 months after treatment completion. Statistical analyses A description of demographic, histologic, and clinical data was performed and presented as mean (SD) for quantitative variables, as frequency tables for qualitative ones. Local recurrence-free survival was defined as relapse occurring in the CTV and locoregional recurrence-free survival as a relapse occurring in the CTVor nodal pelvic recurrence; overall survival was defined as a death from any cause. Local recurrence-free survival was analyzed using the Cox proportional hazard regression model. The effect of DVH parameters on local control was measured in multivariate Cox proportional hazard models adjusted for a selection of
demographic, histologic, and clinical prognostic factors. Wald’s tests on Cox proportional hazard models were then performed to highlight the specific effect of each of these parameters. Prognostic factors including age, tumor size, pelvic lymph nodes, and para-aortic lymph nodes were selected by a stepwise procedure modeling local recurrence-free survival on Cox proportional hazard multivariate models. Note that the FIGO stage was not selected by this procedure as it is strongly statistically linked to the combination of tumor size and lymph node involvement. Threshold doses that maximize the probability of local recurrence were determined by maximum likelihood estimation on Cox proportional hazard models. A p-value #0.05 was considered statistically significant. All calculations were performed with R software version 2.11.0 (R Foundation for Statistical Computing, Vienna, Austria). Results Treatment The pelvic EBRT was delivered with a four-field technique. Median pelvic central dose was 45.5 Gy (2.6 Gy) in 25 fractions. EBRT included para-aortic irradiation for 18 patients (16%) and a parametrial boost ($8 Gy) for 19 of them (17%). Concomitant chemotherapy, mainly weekly cisplatin, was delivered to 90% of patients. BT treatment planning was based on CT scan for 87% of the patients and on MRI for 13%. Dosimetric parameters are listed in Table 2. The mean dose to Point A was 70 Gya/b10 (9.6 Gy). The mean HR-CTV D90 was 73.2 Gya/b10 (11.3 Gy). The mean V85 for HR-CTV was 59% (25). The mean IR-CTV D90 was 61.8 Gya/b10 (6.9 Gy), and the mean V60 for IR-CTV was 86% (20). Table 2 Brachytherapy dosimetric data
Dosimetric parameters
Mean SD
Adjusted hazard of local recurrence p-valuea
Dose to Point A (Gya/b10) TRAK (Gy$cm2) HR-CTV Volume (cm3) D90 (Gya/b10) D100 (Gya/b10) V85 (%) IR-CTV Volume (cm3) D90 (Gya/b10) D100 (Gya/b10) V60 (%)
70 9.6 184 105
0.05 0.001
35 73.2 62.0 59
27 11.3 8.0 25
0.95 0.18 0.28 0.23
99 61.8 54.4 86
56 6.9 4.7 20
0.53 0.39 0.09 0.01
TRAK 5 total reference air kerma; HR-CTV 5 high-risk clinical target volume; IR-CTV 5 intermediate-risk clinical target volume; D90 5 dose received by 90% of target volume; D100 5 minimal target dose; V60 5 percentage of target volume receiving 60 Gy; V85 5 percentage of target volume receiving 85 Gy. a Wald test p-value in Cox proportional hazard model adjusted for age, tumor size, pelvic lymph nodes, and para-aortic lymph nodes.
S. Boyrie et al. / Brachytherapy 13 (2014) 326e331
The median overall treatment time (from the EBRT first session to BT) was 55 days (15). Clinical outcome Median followup was 24.2 months (range, 2.5e39.1); 85% of the patients had a followup time of more than 20 months or until they relapsed. At 24 months, local recurrence-free survival reached 78% (71; 87) and locoregional recurrence-free survival 70% (61; 79). The median delay to local relapse was 10 months (range, 3.7e24.4 months). Overall survival at 24 months was 74% (70; 83). Those results are accurately described elsewhere (4). Dosimetric predictive factors for local control Well established for 2D treatment planning, dose to Point A and total reference air kerma were confirmed to be predictive factors for local control in our study ( p 5 0.001 and p 5 0.001). Patients with a dose to Point A !68 Gya/b10 had a relative risk of local recurrence of 6.1 (95% confidence interval, 1.7e21.7). Concerning 3D DVH parameters, IR-CTV V60 was in our study significantly lower for patients with local relapse than for patients with local tumor control ( p 5 0.013). Patients with an IR-CTV V60 !75% had a relative risk of local recurrence of 3.8 (95% confidence interval, 1.4e11.1) (Fig. 1). HR-CTV D90 and D100 had no significant impact on local control, nor did HR-CTV V85. No significant dependence of local control on doses per pulse for both HRCTV and IR-CTV was found.
Fig. 1. Local recurrence-free survival curves according to IR-CTV V60. IR-CTV 5 intermediate-risk clinical target volume; V60 5 percentage of target volume receiving 60 Gy.
329
Dosimetric study of local relapses Eighteen patients developed a local recurrence. Dosimetric parameters of these patients are listed in Table 3. Mean dose to Point A was 67.3 Gya/b10 (8.8 Gy). The mean HR-CTV D90 was 69.8 Gya/b10 (11.4 Gy), and the mean V85 for HR-CTV was 50% (26). The mean IR-CTV D90 was 59.4 Gya/b10 (7.8 Gy), and the mean V60 for IR-CTV was 75% (26). Ten complete files with local recurrence could be recovered from the different centers involved in the study, including imaging performed at the time of local relapse (patients 2, 5, 6, 7, 8, 12, 13, 16, 17, and 18). When comparing BT treatment planning and imaging performed at the time of local relapse, we found that nine local relapses occurred within the HR-CTV volume. Six patients had target volumes with insufficient dose coverage (mean HR-CTV D90 5 64.2 Gy 7.8 Gy). Among them, 3 patients had a parametrial involvement poorly covered because of using intracavitary BT alone without interstitial needles. For the 3 other patients who presented a local relapse within the HR-CTV, no dosimetric explanation could be given (mean HR-CTV D90 5 85.3 Gy 1.9 Gy). Only 1 patient had an isolated vaginal relapse outside IR-CTV. It is noted that for all these patients, doses to OAR were easily respected.
Discussion Currently, the impact of dosimetry optimization of 3D image-based BT on clinical outcome has been investigated in only a few studies. Dosimetric parameters that were previously reported as predictive factors for local control included doses delivered to the HR-CTV (HR-CTV D100 and particularly HR-CTV D90). Our study is, to our knowledge, the first one to show that IR-CTV coverage is important for tumor control. We have shown in this study that IR-CTV V60 was predictive for local control ( p 5 0.013). Dimopoulos et al (13) have studied the doseeeffect relationship for local control of locally advanced cervical cancer treated with MRI-optimized treatment planning in combination with interstitial BT. His team reported a trend toward significance for local control for the IR-CTV D90, whereas HR-CTV D100 and HR-CTV D90 were clearly significant. Tumor control rates higher than 90% at 3 years could be expected at doses O86 Gya/b10 and 67 Gya/b10 for D90 and D100, respectively. Similar results were reported by Beriwal et al (14) for Pittsburgh with a local control rate of 88% at 2 years using a hybrid MRI/CT approach (mean HR-CTV D90 5 84 Gy) and by Nomden et al (15) for Utrecht achieving local control rates of 93% at 3 years using MRI-guided BT (mean HR-CTV D90 5 83.3 Gy). Delivering such doses to HR-CTV D90 is necessary but not always sufficient to achieve local control. Because the D90 does not reflect the dose to the whole target, lowdose regions within the target may not be apparent. Others have shown that the minimum point dose is also important
330
S. Boyrie et al. / Brachytherapy 13 (2014) 326e331
Table 3 Dosimetric data of patients with local recurrence
Patients 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Mean SD Median (minemax)
FIGO Tumor stage size (mm)
Pelvic lymph HR-CTV TRAK nodes volume (cm3) (Gy.cm2)
Dose to Point A (Gya/b10)
HR-CTV D90 HR-CTV (Gya/b10) V85 (%)
IR-CTV D90 IR-CTV (Gya/b10) V60 (%)
IB1 IIIB IIIB IIB IIIB IB2 IIIB IIB IIA IIB IIIB IB2 IIB IIIA IB2 IB2 IIB IIB
þ þ þ þ þ þ þ þ þ þ þ
56 59 63 63 65 64 70 59 67 64 65 68 67 65 92 84 65 76 67 9 65 (56e92)
51 54 56 58 59 63 63 68 70 70 73 75 77 80 82 84 84 88 70 11 70 (51e88)
na 54 53 52 59 60 54 62 63 68 55 62 45 na 65 53 72 73 59 8 59 (45e73)
40 80 60 30 50 80 na 65 31 na na 55 80 75 60 60 50 48 58 17 60 (30e80)
38 41 67 27 48 36 29 10 34 71 21 17 14 37 106 25 14 15 36 24 32 (10e106)
60 124 111 104 162 132 125 69 148 173 143 164 157 222 320 157 174 225 154 60 152 (60e225)
8 19 24 33 29 34 35 29 43 48 55 62 78 93 70 78 84 87 50 26 45 (8e93)
na 74 69 59 74 95 71 95 96 99 83 29 60 na 99 95 10 100 75 26 78 (10e100)
FIGO 5 Federation of Gynecology and Obstetrics; TRAK 5 total reference air kerma; HR-CTV 5 high-risk clinical target volume; IRCTV 5 intermediate-risk clinical target volume; na 5 nonavailable; D90 5 dose received by 90% of target volume; V60 5 percentage of target volume receiving 60 Gy; V85 5 percentage of target volume receiving 85 Gy; min 5 minimum; max 5 maximum.
to consider for local control. In their study, Schmid et al (16) found a mean minimum point dose to the HR-CTV of 72 Gya/b10 for patients with local recurrences as opposed to 99 Gya/b10 in the matched group of patients with complete local remission. In our study, the HR-CTV dosimetric factors considered here were not found to be predictive of local control. The absence of a doseeresponse relationship between HR-CTV dose-volume parameters and local control in our study can be partially explained by the relatively low doses delivered to the HR-CTV. Indeed, the mean HRCTV D90 was 73.2 Gya/b10, which is lower than in previous studies. There are several possible explanations for this observation. The main aim of the study was to lower the complication rates. So most centers tried to remain within ICRU maximal doses as 3D OARs constraints were not well established at the time of the study. For most centers, 3D planning for gynecologic BT was a new technique. During this learning period, participating centers carefully moved from 2D to 3D planning without necessarily aiming at increasing doses at the same time. Each center was free to optimize treatment planning according to previous procedures as the study started recruiting in January 2005 before GEC-ESTRO recommendations were published. Thus, the study design did not include major dosimetric constraints. A flaw of this multicentric study is heterogeneity of practice, with disparate doses to target volumes and large SDs. But it reflects dose-prescribing practices across the whole country at that period of time. Finally, these low doses may also be partly explained by using CT scans
for delineation of target volumes. Delineation was mostly based on CT scans (87%) as most centers did not have the possibility to perform MRI for dosimetry at the time of BT. MRI is recommended as the standard imaging technique to be used for the initial evaluation of local spread and has been found to be superior to CT in parametrial evaluation and in tumor size assessment (17, 18). CT tumor contours have been found to overestimate the tumor width compared with MRI-based contours (19, 20). This can result in artificially lower D90 and D100 with CT plans compared with MRI-guided dosimetry. Conclusions This multicentric study reports high local control rates with 3D-optimized BT. It is the first study to demonstrate the importance of IR-CTV coverage on local control. Indeed, IR-CTV V60 was identified as a predictive factor for local control. HR-CTV DVHs were not predictive for local control, perhaps, because relatively low doses were delivered to that volume as the main aim of the study was to reduce the risk of severe late complications. The objective of a phase II on going trial is now to deliver higher doses to CTVs as it has been shown to improve dramatically local control even in cases of locally advanced disease. Acknowledgments We wish to thank Astrid de Leeuw for the material support for the calculation of the radiobiologically equivalent
S. Boyrie et al. / Brachytherapy 13 (2014) 326e331
dose of PDR BT. We thank Dr Alain Haddad for his contribution in the translation and editing of this article. The following radiation oncologists and physicists are being acknowledged for their contribution to the STIC PDR study: J. Bonnet, P. Magnenet, N. Allieres, T. Lacornerie, Y. Kirova, J. Coulot, P. Romestaing, M.-P. Sotton, P. Lang, G. Boisserie, S. Racadot, C. Malet, A. Chemin, D.Williaume, J. Bellec, C. Krzisch, K. Peignaux, J.-P. Brenier, A. Cussac, A. Lisbona, M.-H. Baron, O. Goubard, T.D. Nguyen, N. Gaillot, F. Lesaunier, V. Talguen, and I. Barillot, B. Dubray. The study was supported by a grant from the French Minister of Health and Sports : ‘‘STIC PDR: Utilisation de la curietherapie pulsee gynecologique (PDR) avec optimisation de la repartition de la dose et dosimetrie tridimensionnelle’’ (Ministere de la sante et des sports, Programme de soutien aux innovations diagnostiques et therapeutiques co^uteuses). References [1] Han K, Milosevic M, Fyles A, et al. Trends in the utilization of brachytherapy in cervical cancer in the United States. Int J Radiat Oncol Biol Phys 2013;87:111e119. [2] Haie-Meder C, Potter R, Van Limbergen E, et al. Recommendations from Gynaecological (GYN) GEC-ESTRO Working Group (I): Concepts and terms in 3D image based 3D treatment planning in cervix cancer brachytherapy with emphasis on MRI assessment of GTV and CTV. Radiother Oncol 2005;74:235e245. [3] Potter R, Haie-Meder C, Van Limbergen E, et al. Recommendations from gynaecological (GYN) GEC ESTRO working group (II): Concepts and terms in 3D image-based treatment planning in cervix cancer brachytherapy-3D dose volume parameters and aspects of 3D image-based anatomy, radiation physics, radiobiology. Radiother Oncol 2006;78:67e77. [4] Charra-Brunaud C, Harter V, Delannes M, et al. Impact of 3D imagebased PDR brachytherapy on outcome of patients treated for cervix carcinoma in France: Results of the French STIC prospective study. Radiother Oncol 2012;103:305e313. [5] Levitchi M, Charra-Brunaud C, Quetin P, et al. Impact of dosimetric and clinical parameters on clinical side effects in cervix cancer patients treated with 3D pulse-dose-rate intracavitary brachytherapy. Radiother Oncol 2012;103:314e321. [6] Girinsky T, Rey A, Roche B, et al. Overall treatment time in advanced cervical carcinomas: A critical parameter in treatment outcome. Int J Radiat Oncol Biol Phys 1993;27:1051e1056. [7] Lang S, Nulens A, Briot E, et al. Intercomparison of treatment concepts for MR image assisted brachytherapy of cervical carcinoma
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18] [19]
[20]
331
based on GYN GEC-ESTRO recommendations. Radiother Oncol 2006;78:185e193. Van de Kamer JB, De Leeuw AA, Moerland MA, et al. Determining DVH parameters for combined external beam and brachytherapy treatment: 3D biological dose adding for patients with cervical cancer. Radiother Oncol 2010;94:248e253. Fowler JF, Van Limbergen EF. Biological effect of pulsed dose rate brachytherapy with stepping sources if short half-times of repair are present in tissues. Int J Radiat Oncol Biol Phys 1997;37: 877e883. Visser AG, van den Aardweg GJ, Levendag PC. Pulsed dose rate and fractionated high dose rate brachytherapy: Choice of brachytherapy schedules to replace low dose rate treatments. Int J Radiat Oncol Biol Phys 1996;34:497e505. Dale RG, Coles IP, Deehan C, et al. Calculation of integrated biological response in brachytherapy. Int J Radiat Oncol Biol Phys 1997;38: 633e642. Lambin P, Gerbaulet A, Kramar A, et al. Phase III trial comparing two low dose rates in brachytherapy of cervix carcinoma: Report at two years. Int J Radiat Oncol Biol Phys 1993;25:405e412. Dimopoulos JC, Potter R, Lang S, et al. Dose-effect relationship for local control of cervical cancer by magnetic resonance image-guided brachytherapy. Radiother Oncol 2009;93:311e315. Beriwal S, Kannan N, Kim H, et al. Three-dimensional high dose rate intracavitary image-guided brachytherapy for the treatment of cervical cancer using a hybrid magnetic resonance imaging/computed tomography approach: Feasibility and early results. Clin Oncol (R Coll Radiol) 2011;23:685e690. Nomden CN, de Leeuw AA, Roesnink JM, et al. Clinical outcome and dosimetric parameters of chemo-radiation including MRI guided adaptative brachytherapy with tandem-ovoid applicators for cervical cancer patients: A single institution experience. Radiother Oncol 2013;107:69e74. Schmid MP, Kirisits C, Nesvacil N, et al. Local recurrences in cervical cancer patients in the setting of image-guided brachytherapy: A comparison of spatial dose distribution within a matched-pair analysis. Radiother Oncol 2011;100:468e472. Kim SH, Choi BI, Han JK, et al. Preoperative staging of uterine cervical carcinoma: Comparison of CT and MRI in 99 patients. J Comput Assist Tomogr 1993;17:633e640. Kim SH, Choi BI, Lee HP, et al. Uterine cervical carcinoma: Comparison of CT and MR findings. Radiology 1990;175:45e51. Viswanathan AN, Dimopoulos J, Kirisits C, et al. Computed tomography versus magnetic resonance imaging-based contouring in cervical cancer brachytherapy: Results of a prospective trial and preliminary guidelines for standardized contours. Int J Radiat Oncol Biol Phys 2007;68:491e498. Hegazy N, P€otter R, Kirisits C, et al. High-risk clinical target volume delineation in CT-guided cervical cancer brachytherapy: Impact of information from FIGO stage with or without systematic inclusion of 3D documentation of clinical gynecological examination. Acta Oncol 2013;53:1345e1352.