|Ahead of print publication
Comparison of volume doses from conventional two-dimensional brachytherapy with corresponding doses from three-dimensional magnetic resonance imaging-based brachytherapy in carcinoma cervix
Susan Mathews1, M Biju Azariah1, Seetha Mohandas2, Sharika V Menon3, Preethi George4, PG Jayaprakash2
1 Department of Radiation Oncology, Regional Cancer Centre, Thiruvananthapuram, Kerala, India
2 Department of Clinical Oncology, SUT Hospital, Thiruvananthapuram, Kerala, India
3 Department of Radiation Physics, Regional Cancer Centre, Thiruvananthapuram, Kerala, India
4 Department of Bio Statistics, Regional Cancer Centre, Thiruvananthapuram, Kerala, India
M Biju Azariah,
Department of Radiation Oncology, Regional Cancer Centre, Thiruvananthapuram - 695 011, Kerala
Source of Support: None, Conflict of Interest: None
Purpose: The purpose of this study was to evaluate the doses delivered to the brachytherapy (BT) target volume and organs at risk from two-dimensional X-ray-based plans on magnetic resonance imaging (MRI) and to compare these doses with the corresponding doses from the image-based optimized plans.
Materials and Methods: Twenty patients with cervical cancer treated with chemoradiation and BT were included in this study. All patients had two sets of treatment plans generated for the first fraction of BT. Volume doses resulting from MRI-based optimized plans were compared with the corresponding doses from standard “Point A” prescription plans.
Results: There was statistically significant difference between the two planning modalities for the mean high-risk clinical target volume (HRCTV) D90 doses (P = 0.0014) although mean D2cc of bladder (P = 0.1667) and rectum (P = 0.051) was not different. Standard plans with a prescription dose of 7 Gy to Point A delivered a mean HRCTV D90 of 10.07 Gy in patients with no gross residual disease at the time of BT, which was very similar to the mean dose from MR-based plans (MRI 10.02 Gy and standard 10.07 Gy). The only factor seen affecting dose distribution in this group was the applicator geometry. Standard plans failed to deliver HRCTV D90 doses of >8.5 Gy in all patients with gross residual disease. The doses were <7.00 Gy to the HRCTV in three patients who had maximum residual diseases at the time of BT.
Conclusion: Conventional X-ray-based plans with moderate Point A doses deliver HRCTV D90 comparable to MRI-based plans in patients with no residual disease, and centrally placed residual disease, provided proper applicator placement and ideal geometry can be ensured. Soft-tissue image-based BT dose optimization ought to be considered in all patients with gross residual disease at the time of brachytherapy.
Keywords: Carcinoma cervix, dosimetry, image-based brachytherapy
|How to cite this URL:|
Mathews S, Azariah M B, Mohandas S, Menon SV, George P, Jayaprakash P G. Comparison of volume doses from conventional two-dimensional brachytherapy with corresponding doses from three-dimensional magnetic resonance imaging-based brachytherapy in carcinoma cervix. J Can Res Ther [Epub ahead of print] [cited 2019 May 19]. Available from: http://www.cancerjournal.net/preprintarticle.asp?id=257733
| > Introduction|| |
Brachytherapy for the majority of patients in India and in the developing world continues to be planned using two-dimensional (2D) technique., The results of treating cervical cancers with brachytherapy (BT) by using conventional prescription methods have been good, although we do not have sufficient three-dimensional (3D) dosimetric data for the tumor or normal tissues. With the publication of results from centers across the world, showing improved outcomes and reduced toxicity with the use of image-guided adaptive BT (IGABT) for cervical carcinoma, IGABT has become the gold standard in gynecologic BT and the benchmark for dose optimization.,, However, in the developing world, there is an acknowledged shortage of basic radiation oncology equipment and workforce, making IGABT far from realistic, chiefly because of additional resources needed and the technical complexities involved. With 3D BT, we now have greater insight into the dose-volume-outcome parameters which afford better understanding of conventional BT planning and traditional dose prescription. Conventionally, our cumulative Point A doses have always been in the 70–75 Gy dose range which is lower than the recommended conventional dose prescription for Point A of 75–85 Gy Understanding the doses received by critical volumes from conventional 2D BT would help us understand 2D dosimetry objectively and estimate the benefits of soft-tissue imaging information in BT treatment planning. This information can be used for developing strategies to allocate and manage the existing resources in limited resource environments.
This study evaluates 2D BT dosimetry in terms of volume doses delivered and attempts to understand how and where soft-tissue imaging may improve it.
Purpose of the study
The study aims to compare the doses delivered to the target volumes and normal structures from conventional 2D X-ray-based BT planning with the corresponding doses from MRI-based optimized plan to evaluate 2D dosimetry.
| > Materials and Methods|| |
Twenty patients with carcinoma of the cervix, treated with curative-intent external beam radiotherapy (EBRT) with concurrent chemotherapy and image-based BT, were included in this study. All patients had MRI performed in addition to the planning computed tomography (CT) images acquired after the first BT application, and MR-optimized plans were generated based on the Groupe Européan de Curie Thérapie–European Society of Radiation Oncology (GEC-ESTRO) guidelines., Standard “Point A” prescription plans were also generated on the MR planning images which had the volumes contoured on it, and the resultant volume doses from standard plans were recorded and compared.
Pretreatment patient characteristics
The mean age of the patients was 48 years (range 33–77 years). Twelve patients had bulky tumors measuring >5 cm at presentation. FIGO Stage at presentation was as follows: IB-7, IIB-8, IIIB-4, and one patient with Stage IVA (bladder infiltration) disease. Pretreatment tumor volumes ranged between 11.2 and 265.3 cc, with an average of 96.96 cc.
MR-compatible tandem and ovoid applicators (Nucletron®) were used for each insertion. Tandems typically reached the end of the uterine canal.
High-resolution T2-weighted axial, coronal, and sagittal sequences were acquired and transferred to the Oncentra BT treatment planning system version 4.3, (Nucletron an Elekta company, Elekta AB, Stockholm, Sweden). The gross tumor volume at BT, high-risk-clinical target volume (HR-CTV), and organs at risk (OARs) (bladder, rectum, and sigmoid) were contoured on the acquired images as per the GEC-ESTRO recommendations.
MRIs were acquired as per the GEC–ESTRO guidelines. The HR-CTV and OAR (rectum and bladder) were contoured on MRI following the recommendations of GEC–ESTRO.
Magnetic resonance plan
Theoretical MRI-based plans were generated. The aim was to deliver the highest possible dose to the HRCTV with respect to OARs.
Dose prescription and constraints
The planning aim with MRI-based plan was to deliver a minimum of 80–85 Gy10 to 90% of the HRCTV (D90) in 2 Gy equivalent doses (EQD2) applying the linear quadratic model with an α/β ratio of 10 Gy for tumor (~8.7 Gy/fx for 3 fx high dose rate (HDR) BT) and α/β ratio of 3 Gy for OAR, applying a half-time repair of 1.5 h. To deliver a D90 of >80 Gy10 to the HRCTV, a prescription dose of 8–8.5 Gy was chosen. Dose constraints applied to OARs were a maximum dose of 75 Gy3(~6 Gy/fx) to the maximally exposed D2cc of the rectum and 90 Gy3 (~7.5 Gy/fx) to the bladder (Dose constraints adapted from Subir Nag HDR BT dosimetry data published in 1998. The dose prescription was as for patients typically receiving 45 Gy EBRT and three fractions of BT).
The applicators were reconstructed manually on both the CT study set (used for treatment plan generation) and the corresponding MRIs. The dwell positions corresponding to traditional Manchester loading pattern were activated and the plan was then normalized to Point A with a prescription dose of 7 Gy. Typically, patients receive three treatments at weekly intervals. The doses received by the different volumes contoured on MRI from standard planning were then documented. Standard plans were generated for all patients on the MR plan images according to our long-standing protocol, using standard Manchester loading. A dose of 7 Gy was prescribed to Point A. Conventionally, our BT dose after 40/45 Gy EBRT to the pelvis has been 7 Gy to Point A delivered at weekly interval for 3 weeks. This dose was arrived at based on our institutional experience, balancing outcome and toxicity to OAR.
Residual disease at the time of BT was documented based on clinical assessment. Cervical dimensions were measured on the MRIs and documented in terms of height, width, and thickness. For dosimetric interpretations, patients were grouped on the basis of their residual disease status taking the maximum measurement in any direction as follows: Group 1: patients with superficial/<1 cm minimal or no gross residual disease; Group 2: patients with residual disease measuring 1–3 cm; and Group 3: patients with a bulky cervix measuring >3 cm in any direction. Doses received by target and normal structures were analyzed separately for each modality for these three groups.
Data were summarized as means, standard deviations (SDs), and ranges. The planning methods were compared using two-way analysis of variance. Two methods were compared using Student's t-tests, and then nonparametric Kruskal–Wallis test was also performed as the sample size was small. Post hoc test was done to identify the modality which showed significant difference. P < 0.05 was considered statistically significant. Statistical evaluation was performed using SPSS statistical software (version 11.0 for Windows, SPSS Inc., Chicago, IL, USA).
| > Results|| |
The mean HRCTV volume for the entire cohort was 28.2 cc (SD – 14.64, range: 16.25–75.53 cc and median 23.2 cc).
From among the twenty patients, there were eight patients in Group 1, five patients with no residual disease, and three with minimal residual disease. Their HRCTV volumes ranged from 16.25 to 23.47 cc with a mean HRCTV of 19.38 cc.
Six patients had residual disease measuring 1–3 cm and were included in Group 2. The mean HRCTV was 23.17 cc (SD – 3.68, range: 18.64–27.71, and median 23 cc).
There were six patients with residual tumors measuring >3 cm, having a mean HRCTV of 45 cc (SD – 17.29, range: 29.7–75.53, and median of 40.49 cc), included in Group 3.
Three patients in this group had residual disease asymmetrically located, predominantly in the posterior lip. Clinically maximum residual disease recorded was 5 cm × 4 cm residue almost entirely in the posterior lip.
Two patients in Group 1 had nonideal applications as seen on postprocedural images. The tandem was not in the endometrial cavity but had penetrated the myometrium beyond the cervix but did not perforate the uterus. The HRCTVs were 16.25 and 22.92 cc, respectively. The HRCTV D90 resulting from nonideal insertions was documented. The maximum HRCTV of 75.53 cc was for a patient who had cystic lesions in the cervix uniformly expanding the volume. In the patient with bladder infiltrative disease at presentation, the initial tumor-bearing region was replaced by an ulcer at the time of BT, the wall of which was included in the HRCTV.
Standard plan dosimetry
Standard plans, with a prescription dose of 7 Gy to Point A, resulted in mean HRCTV D90 of 8.65 Gy overall. The HRCTV D90 doses ranged from a minimum of 4.69 Gy to a maximum of 11.64 Gy. The HRCTV D90 was minimum (4.69 Gy) in the patient with maximum residual tumor which was seen extending 3.5 cm from the tandem posteriorly.
Standard plan dosimetry was seen to deliver doses >8.7 Gy to HRCTV D90 (per fraction dose for 3 HDR fraction BT which would deliver a minimum of 85 Gy10 to 90% of the HRCTV (D90) in 2 Gy EQDs) in all but three patients (11 out of 20) with no residual disease or <3 cm residual disease (Groups 1 and 2). Of the three patients who did not achieve the planning goal, two patients had nonideal insertions. The HRCTV D90 doses were suboptimal at 7.71 Gy and 7.65 Gy for 16.25 and 22.92 cc HRCTVs, respectively. In the third patient, the HRCTV D90 was 8.57 Gy, just under the planning aim. Here, the residual disease, though minimal, was more to one side, the asymmetry accounting for reduction in HRCTV D90 dose.
Standard plan dosimetry failed to deliver HRCTV D90 doses of >8.5 Gy in all the six patients with gross residual disease. The doses were <7.00 Gy to the HRCTV in all the three patients with maximum residual disease at the time of BT. This indicates that in Group 3 patients the tumor extended well beyond Point A where the dose was prescribed. The maximum dose that could be delivered with standard plan in this group was 8.27 Gy, for a volume of 40.46 cm3, where the residual disease was essentially endocervical. The HRCTV volumes and corresponding D90 doses from standard plan were as follows: 75.53 cc–5.48 Gy, 53.34 cc–4.69 Gy, 40.46 cc–8.27 Gy, 40.53 cc–6.49 Gy, 30.24 cc–8.11 Gy, and 29.71 cc–7.36 Gy.
Organs at risk doses
With standard plan, the bladder tolerance exceeded in patients with the smallest HRCTVs, particularly where the volumes were <20 cc. Thus, in four patients, with no residual disease, the D2cc bladder doses exceeded 7.5 Gy. In the patient with Stage IVA disease (bladder) at presentation, the standard plan resulted in unacceptably high bladder D2cc dose of 11.32 Gy. The D2cc bladder dose was >9 Gy in one of the patients with nonideal application. The D2cc rectal dose was well within the tolerance limit in all but one patient where the dose was 6.6 Gy.
Comparison of methods
Overall, there was statistically significant difference observed between the two planning modalities for the mean HRCTV D90 doses (P = 0.0014). There were no statistically significant differences for mean doses to the OAR for the modalities [Table 1].
|Table 1: Comparison of mean doses for HRCTV D90 and D2 cm3 Bladder and Rectum|
Click here to view
There were variable differences between the two planning modalities for the different residual disease groups. For patients with no gross residual disease at the time of BT, the mean doses were similar in both plans (P = 0.846) [Table 2]. The differences in HRCTV coverage in Group 2 patients with residual disease measuring 1–3 cm were statistically significant (P = 0.044). However, with standard plan, the mean HRCTV D90 in this group was 9.46 Gy which is well beyond the planning goal.
|Table 2: Comparisons of the mean HRCTV D90 across the imaging modalities for the different groups based on residual disease|
Click here to view
In Group 3 patients, with 3 cm or more residual disease, there were significant differences between the mean HRCTV D90 delivered between the standard plans and MR-based plans (P < 0.0001). With MR plan, the optimization goal could be achieved in four out of the six Group 3 patients. In the patient with gross asymmetry (residual tumor seen extending 3.5 cm posteriorly from the tandem), delivery of curative BT dose was considered impossible with standard applicators. The maximum HRCTV D90 achievable in this patient was 7 Gy with MRI-based optimization, limiting D2cc bladder dose to 7.57 Gy and D2cc rectal dose to 6.09 Gy. Comparisons of the mean HRCTV D90 across the imaging modalities for the different groups based on residual disease are represented in [Table 2] and [Table 3].
|Table 3: Comparison of doses between imaging modalities with respect to HRCTV volume|
Click here to view
| > Discussion|| |
The Manchester System of Dosimetry, the last major revision of which came in 1953, is still the most widely used system for gynecologic BT planning. With its strict set of rules and dose points defined on the basis of clinical outcome correlation, it has stood the test of time and given gynecologic intracavitary BT a strong foundation and a reproducible framework. Soft-tissue imaging information has undoubtedly benefited all intracavitary planning by enabling dose to be optimized to a target volume. We used a moderate prescription dose of 7 Gy to Point A for standard plans. Interestingly, Point A doses from recent image-based studies attest to the reasonableness of this dose for small HRCTVs. In a report on early outcomes of 128 patients receiving MRI-guided BT, Gill et al. found smaller HRCTV volumes that resulted in lower Point A doses of <75 Gy10. A prescription of 7 Gy in our study resulted in a total dose of 74 Gy10 to Point A.
For patients with no gross residual disease at the time of BT, the mean HRCTV doses were similar for both planning modalities. Excluding the two patients for whom the tandem was not in the central canal, the mean HRCTV D90 values were almost uniform across the modalities (MRI 10.02 Gy, standard 10.07 Gy). The only factor affecting dose distribution in this group was the applicator geometry. Of the 20 patients, 14 patients had no residual or minimal residual disease at BT and this study has shown that standard planning delivers adequate HRCTV doses in these patients.
We detected two nonideal insertions in this study. For patients with nonideal applications, volume asymmetry resulting from tandem being out of the central canal resulted in underdosing of a portion of HRCTV with Point A prescriptions. This compromise in HRCTV dose could not be improved with sophisticated planning. In this study, the maximum HRCTV doses that could be delivered using gold standard MRI optimization in these two patients were 8.00 and 8.05 Gy, respectively, respecting OAR tolerances. Ultrasound image guidance to ensure ideal applicator geometry would guarantee the success of standard planning using 2D X-rays in patients with no/minimal residual disease at BT.,,
For patients with minimal residual disease and disease located centrally, the mean HRCTV D90 doses were MRI 10.39 Gy and standard planning 9.46 Gy. While the difference was statistically significant (P = 0.044), the standard plan delivered mean HRCTV D90 dose >9.00 Gy, indicating that the difference may not be clinically significant. Central residual disease (endocervical residue, corpus involvement) was adequately covered with the standard plan. Again, volume asymmetry was critical in deciding the HRCTV dose with standard planning. When there was asymmetrically located disease even in small-volume residual disease, the coverage achieved with standard plans deteriorated.
There were significant differences in the mean HRCTV D90 doses delivered with standard plans (6.73 Gy) and MRI-based plans (8.83 Gy) in patients with gross residual disease measuring >3 cm (P < 0.0009). For standard plans, the location of the disease with respect to the central tandem affected the HRCTV doses more than the bulk of residual disease. In two patients with gross residual disease symmetrically located about the central tandem, increasing the Point A prescription dose would have resulted in the aimed dose with standard plan. Asymmetric tumor extending 3 cm or more to any one side of the tandem resulted in significant underdosing of the HRCTV. Often, this extent of asymmetry requires radiation delivery via interstitial techniques for curative doses to be delivered without exceeding OAR tolerance limits. However, some of the limitations of standard applicators can be overcome through the use of soft-tissue imaging which enables us to guide optimization and spare doses to OAR.
Organs at risk doses
The D2cc bladder doses exceeded 7.5 Gy with standard plan in four patients with no residual disease. The D2cc bladder doses were >9 Gy in two patients. It was as high as 11.32 Gy in the patient with bladder infiltrative disease where the tissue loss at the region of cervix resulted in the intrauterine tandem being very close to bladder wall. The D2cc bladder dose was >9 Gy in one of the patients with nonideal application too. The D2cc rectal doses were well within the tolerance limit in all but one patient where the dose was 6.6 Gy. The lower rectal doses in all patients can be attributed to the use of rectal retractor which is part of the MR-compatible applicator set used.
This study has shown that the extent of dose optimization possible is limited by applicator availability (interstitial capabilities), individual patient anatomy, and disease biology. Patient selection based on residual disease status could be performed at first BT application based on clinical assessment and ultrasound imaging information. Patients with asymmetric and bulky residual disease should be considered for treatment utilizing applicators with interstitial capabilities. Planning with these is currently best served by MRI and ought to be practiced where facilities exist.
| > Conclusion|| |
This study suggests that conventional X-ray-based planning with a moderate Point A prescription dose delivers comparable HRCTV D90 doses in patients with no residual or minimal residual disease, provided proper applicator placement and ideal geometry can be ensured. Tumor asymmetry matters more than tumor bulk in conventional planning as bulky centrally placed residual disease could still be adequately treated with standard plans.
Patient selection based on residual disease status at BT would enable us to allocate and manage resources and effectively deliver the benefits of image guidance to all patients. Significant improvements to current treatments based on standard 2D X-ray plans are possible by incorporating ultrasound imaging for guiding applicator placements and improving technical accuracy.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| > References|| |
Banerjee S, Mahantshetty U, Shrivastava S. Brachytherapy in India – A long road ahead. J Contemp Brachytherapy 2014;6:331-5.
Ali I, Wani WA, Saleem K. Cancer scenario in India with future perspectives. Cancer Ther 2011;8:56-70.
Pötter R, Dimopoulos J, Georg P, Lang S, Waldhäusl C, Wachter-Gerstner N, et al.
Clinical impact of MRI assisted dose volume adaptation and dose escalation in brachytherapy of locally advanced cervix cancer. Radiother Oncol 2007;83:148-55.
Pötter R, Georg P, Dimopoulos JC, Grimm M, Berger D, Nesvacil N, et al.
Clinical outcome of protocol based image (MRI) guided adaptive brachytherapy combined with 3D conformal radiotherapy with or without chemotherapy in patients with locally advanced cervical cancer. Radiother Oncol 2011;100:116-23.
Rijkmans EC, Nout RA, Rutten IH, Ketelaars M, Neelis KJ, Laman MS, et al.
Improved survival of patients with cervical cancer treated with image-guided brachytherapy compared with conventional brachytherapy. Gynecol Oncol 2014;135:231-8.
Datta NR, Samiei M, Bodis S. Radiation therapy infrastructure and human resources in low- and middle-income countries: Present status and projections for 2020. Int J Radiat Oncol Biol Phys 2014;89:448-57.
Dimopoulos JC, Petrow P, Tanderup K, Petric P, Berger D, Kirisits C, et al.
Recommendations from gynaecological (GYN) GEC-ESTRO working group (IV): Basic principles and parameters for MR imaging within the frame of image based adaptive cervix cancer brachytherapy. Radiother Oncol 2012;103:113-22.
Haie-Meder C, Pötter R, Van Limbergen E, Briot E, De Brabandere M, Dimopoulos J, 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;74:235-45.
Nag S, Gupta N. A simple method of obtaining equivalent doses for use in HDR brachytherapy. Int J Radiat Oncol Biol Phys 2000;46:507-13.
Tod M, Meredith WJ. Treatment of cancer of the cervix uteri, a revised Manchester method. Br J Radiol 1953;26:252-7.
Viswanathan AN, Erickson BA. Seeing is saving: The benefit of 3D imaging in gynecologic brachytherapy. Gynecol Oncol 2015;138:207-15.
Gill BS, Kim H, Houser CJ, Kelley JL, Sukumvanich P, Edwards RP, et al.
MRI-guided high-dose-rate intracavitary brachytherapy for treatment of cervical cancer: The university of Pittsburgh experience. Int J Radiat Oncol Biol Phys 2015;91:540-7.
Granai CO, Doherty F, Allee P, Ball HG, Madoc-Jones H, Curry SL, et al.
Ultrasound for diagnosing and preventing malplacement of intrauterine tandems. Obstet Gynecol 1990;75:110-3.
Davidson MT, Yuen J, D'Souza DP, Radwan JS, Hammond JA, Batchelar DL, et al.
Optimization of high-dose-rate cervix brachytherapy applicator placement: The benefits of intraoperative ultrasound guidance. Brachytherapy 2008;7:248-53.
Small W Jr., Strauss JB, Hwang CS, Cohen L, Lurain J. Should uterine tandem applicators ever be placed without ultrasound guidance? No: A brief report and review of the literature. Int J Gynecol Cancer 2011;21:941-4.
van Dyk S, Schneider M, Kondalsamy-Chennakesavan S, Bernshaw D, Narayan K. Ultrasound use in gynecologic brachytherapy: Time to focus the beam. Brachytherapy 2015;14:390-400.
[Table 1], [Table 2], [Table 3]