- Email: [email protected]
Volumetric modulated arc radiotherapy for carcinomas of the oro-pharynx, hypo-pharynx and larynx: A treatment planning comparison with fixed field IMRT
Radiotherapy and Oncology 92 (2009) 111–117
Contents lists available at ScienceDirect
Radiotherapy and Oncology journal homepage: www.thegreenjournal.com
Head and neck IMRT
Volumetric modulated arc radiotherapy for carcinomas of the oro-pharynx, hypo-pharynx and larynx: A treatment planning comparison with fixed field IMRT Eugenio Vanetti a, Alessandro Clivio a, Giorgia Nicolini a, Antonella Fogliata a, Sarbani Ghosh-Laskar b, Jai Prakash Agarwal b, Ritu Raj Upreti b, Ashwini Budrukkar b, Vedang Murthy b, Deepak Dattatray Deshpande b, Shyam Kishore Shrivastava b, Ketayun Ardeshir Dinshaw b, Luca Cozzi a,* a b
Oncology Institute of Southern Switzerland, Radiation Oncology Department, Bellinzona, Switzerland Departments of Radiation Oncology & Medical Physics, Tata Memorial Hospital, Mumbai, India
a r t i c l e
i n f o
Article history: Received 2 September 2008 Received in revised form 23 December 2008 Accepted 26 December 2008 Available online 20 January 2009 Keywords: RapidArc IMRT Head and neck radiation therapy
a b s t r a c t Purpose: A planning study was performed to evaluate the performance of volumetric modulated arc radiotherapy on head and neck cancer patients. Conventional fixed field IMRT was used as a benchmark. Methods and materials: CT datasets of 29 patients with squamous cell carcinoma of the oro-pharynx, hypo-pharynx and larynx were included. Plans for fixed beam IMRT, single (RA1) and double (RA2) modulated arcs with the RapidArc technique were optimised. Dose prescription was set to 66 Gy to the primary tumour (at 2.2 Gy/fraction), 60 Gy to intermediate-risk nodes and 54 Gy to low-risk nodal levels. The planning objectives for PTV were minimum dose >95%, and maximum dose <107%. Maximum dose to spinal cord was limited to 46 Gy, maximum to brain stem to 50 Gy. For parotids, mean dose <26 Gy (or median <30 Gy) was assumed as the objective. The MU and delivery time were scored to measure expected treatment efficiency. Results: Target coverage and homogeneity results improved with RA2 plans compared to both RA1 and IMRT. All the techniques fulfilled the objectives on maximum dose, while small deviations were observed on minimum dose for PTV. The conformity index (CI95%) was 1.7 ± 0.2 for all the three techniques. RA2 allowed a reduction of D2% to spinal cord of 3 Gy compared to IMRT (RA1 D2% increased it of 1 Gy). On brain stem, D2% was reduced from 12 Gy (RA1 vs. IMRT) to 13.5 Gy (RA2 vs. IMRT). The mean dose to ipsi-lateral parotids was reduced from 40 Gy (IMRT) to 36.2 Gy (RA1) and 34.4 Gy (RA2). The mean dose to the contra-lateral gland ranged from 32.6 Gy (IMRT) to 30.9 Gy (RA1) and 28.2 Gy (RA2). Conclusion: RapidArc was investigated for head and neck cancer. RA1 and RA2 showed some improvements in organs at risk and healthy tissue sparing, while only RA2 offered improved target coverage with respect to conventional IMRT. Ó 2009 Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology 92 (2009) 111–117
The aim of the present study was to investigate the potential clinical role for head and neck cancer patients of RapidArc, the novel radiation treatment technique (Varian Medical Systems), which is based on volumetric intensity-modulated arc delivery, as opposed to intensity modulation which uses fixed gantry beams. RapidArc falls into the category of intensity modulation therapy with arcs (IMAT) [1–4]. The Yu’s group established the benefit of using multiple modulated arcs for complex cases [5,6]. The Ghent group applied IMAT techniques with multiple non-coplanar beams to pelvic treatments [7,8] proving equivalent or superior target coverage and improved sparing of OARs compared to conventional conformal treatments.
* Corresponding author. Address: Oncology Institute of Southern Switzerland, Radiation Oncology Department, Medical Physics Unit, 6504 Bellinzona, Switzerland. E-mail address: [email protected] (L. Cozzi). 0167-8140/$ - see front matter Ó 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.radonc.2008.12.008
RapidArc is a technique based on an investigation from K. Otto [9] and it aims to (i) improve OARs and healthy tissue sparing compared to other solutions; (ii) maintain or improve the same degree of target coverage; and (iii) reduce beam-on time per fraction. Faster treatments could have a clinical impact on patients in terms of comfort on couch, immobility and minimisation of internal organ displacement. It could also allow more time for imaging procedures allowing, in perspective, routine application of adaptive treatment strategies when changes are observed due to response to radiation as in patients with head and neck cancer. IMRT in head and neck cancer patients has been largely investigated at both planning and clinical levels. Excellent reviews for treatment outcome and major toxicity patterns can be found in Gregoire et al. [10], Lee et al. [11] and Popovtzer et al. [12]. On the toxicity side, besides the major attention given to spinal cord and brain stem (with toxicity thresholds set in the proximity of 45–50 Gy for the first and at 50 Gy for the second), it is generally
112
Volumetric arc modulation for head and neck radiation therapy
known that, for parotids, mean doses inferior to 25–30 Gy correlate well with substantial recovery of function within two years [13] (higher thresholds were observed for sub-mandibular glands in the range of 39 Gy [14]). To reduce dysphagia [15], sparing of constrictors (with mean dose below 60 Gy) highly correlated with improved swallowing, laryngeal elevation and epiglottic inversion. To manage ototoxicity [16], the mean dose to cochlea proved to be a highly significant factor with some threshold effect in the order of 60 Gy between high and low risks of auditory defects. Volumetric IMAT has already been investigated for prostate, small brain tumours and cervix uteri cancer [17–19], i.e. on relatively simple clinical cases; head and neck is an ideal advanced benchmark for assessment of its conformal avoidance capabilities since the anatomical features of this location require highly sophisticated techniques to ensure adequate treatments. The present study was initiated as a side investigation in the framework of a larger Phase II trial activated at Tata Memorial Centre (TMC) in Mumbai to investigate the role of IMRT vs 3D conformal radiotherapy in squamous cell carcinoma (SCC) patients at AJCC stages T1-3, N0-2b, and M0.
Material and methods Patient selection and planning objectives CT data (3 mm slice thickness) for a group of twenty-nine consecutively treated patients from the TMC protocol were selected for the purposes of planning. These patients had been randomized and treated on the IMRT arm of the protocol. The RapidArc planning was carried out on the same dataset at the Oncology Institute of Southern Switzerland. Fourteen tumours were localised in the oro-pharynx district, eight in the hypo-pharynx and seven in the larynx. Fourteen patients presented T3 stage, twelve T2 stage and two T1 stage. Nodal involvement was mostly N0 (16), N1 (6) and N2 (7) stages; all patients presented no distant metastasis (M0). The main organs at risk (OAR) considered for all patients were ipsi- and contra-lateral parotids, spinal cord and brain stem. For some of the patients, additional organs were outlined by radiation oncologists depending on the indication including oral cavity, oesophageal constrictors, base of the tongue, cochlea, mandible, and vocal apparatus. These additional OARs were defined only for those patients where potential sparing was possible due to the target conformation. The healthy tissue was defined as the patient’s volume covered by the CT scan minus the envelope of the various target volumes (PTV). Various PTVs were defined from the respective clinical target volumes (CTVs) adding 7 mm margins with 3D expansion. Dose prescription was set to 66 Gy at 2.2 Gy/fraction to the PTV including the primary tumour and the lymph-nodal metastases (PTV66 – high risk). Two additional elective PTVs were defined to be irradiated at 60 Gy (intermediate risk) and 54 Gy (low risk, sub-clinical disease) in the nodal regions (PTV60 and PTV54). All volumes were to be simultaneously treated according to the simultaneous integrated boost (SIB) approach. Due to target definition criteria, PTVs did not overlap and were not mutually included. All plans were normalised to the mean dose of PTV66. The mean volume of target volumes was PTV66: 342 ± 182 cm3, PTV60: 107 ± 81 cm3, and PTV54: 95 ± 54 cm3 for the global group. For all the PTVs, plans aimed to achieve minimum dose larger than 95% of the prescribed dose and a maximum lower than 107%. For the spinal cord, a maximum dose of 46 Gy was allowed. For the brain stem, the limit was set to 50 Gy. In the case of parotids, the planning objectives aimed to keep the mean dose below 26 Gy (or
D50% < 30 Gy). For other organs at risk, the planning strategy was to minimise their involvement but no specific constraint was set. Planning techniques Three sets of plans were compared in this study, all designed on the Varian Eclipse treatment planning system (TPS) with 6 MV photon beams from a Varian Clinac equipped with a Millennium Multileaf Collimator (MLC) with 120 leaves (spatial resolution of 5 mm at isocentre for the central 20 cm and of 10 mm in the outer 2 10 cm, maximum leaf speed of 2.5 cm/s and leaf transmission of 1.8%). Plans for RapidArc were optimised selecting a maximum DR of 600 MU/ min, and a fixed DR of 300 MU/min was selected for IMRT. The Anisotropic Analytical Algorithm (AAA, version 8.6.02) photon dose calculation algorithm was used for all cases [20–23]. The dose calculation grid was set to 2.5 mm. RapidArc optimisation was performed with version 8.6.05. Details for each planning method are as follows IMRT Reference plans were computed by selecting the ‘conventional’ intensity modulation approach as a benchmark, with fixed gantry and intensity-modulated beams delivering the dose by means of the sliding window approach [24–26]. Plans were individually optimised using seven or nine coplanar fields. The modulation fluences used in the present analysis are equivalent to the fluences approved for patient treatment, while final dose calculation was performed using the AAA, including heterogeneity management, instead of the Pencil Beam (used at TMC) for coherence with the RapidArc condition. MUs were kept fixed at the value from original calculation. It should be noted that in Eclipse, the optimisation process to generate the fluences and the final dose calculation are completely disentangled and therefore the determination of the optimal fluence is in no way influenced by the algorithm adopted afterwards. RapidArc (RA) To achieve the desired level of modulation required, the instantaneous dose rate (DR), MLC leaf positions and the gantry rotational speed were continuously varied by the RapidArc optimiser. To minimise the contribution of tongue and groove effect, the collimator rotation in RapidArc was kept fixed to a value different from zero. In the present study, the collimator was rotated to 40°. Details of the RapidArc process can be found in [9,15]. Two sets of plans were optimised and analysed. RA1 consisting of a single 360° rotation and RA2 consisting of two coplanar arcs of 360° were optimised simultaneously, to be delivered with opposite rotation (clock- and counter clock-wise). The application of two coplanar arcs aims to increase the modulation factor during optimisation. In fact, since each individual arc is limited to a sequence of 177 control points (i.e. ‘’elementary’’ fields), the application of two independent arcs, simultaneously optimised, could allow the optimiser to achieve higher target homogeneity and lower OARs involvement at the same time, as seen in other IMAT applications [5–8]. In the previous investigations on RapidArc [17,18], the relative simplicity of cases did not require investigation of this feature of the optimiser. Evaluation tools Quantitative evaluation of plans was performed by means of standard Dose-Volume Histogram (DVH). For PTV, the values of D98% and D2% (dose received by the 98%, and 2% of the volume) were defined as metrics for minimum and maximum doses in association to V95% V107% (the volume receiving at least 95% or at most
E. Vanetti et al. / Radiotherapy and Oncology 92 (2009) 111–117
107% of the prescribed dose). The target homogeneity was expressed by D5% D95% (difference between the dose covering 5% and 95% of the PTV). The degree of conformality of the plans was measured with a Conformity Index, CI95%, which is defined as the ratio between the patient volume receiving at least 95% of the prescribed dose and the volume of the PTV. For OARs, the analysis included the mean dose, the maximum dose expressed as D2% and a set of appropriate VX and DY values. For Healthy Tissue, the integral dose, ‘‘DoseInt”, is defined as the integral of the absorbed dose extended to over all voxels excluding those within the target volume (DoseInt dimensions are Gy cm3). This was reported together with the observed mean dose and some representative VxGy value. The average cumulative DVH for PTV, OARs and healthy tissue were built from the individual DVHs obtained by averaging the corresponding volumes at each dose bin (0.01 Gy in this case). To appraise the difference between the techniques, the paired, twotailed Student’s t-test was applied. Data were considered statistically significant for p < 0.05. Results Dose distributions are shown for one example in Fig. 1 for axial sagittal and coronal views. Fig. 2 shows the average DVH for all the PTVs comparing the three techniques for the entire patients cohort. Fig. 3 reports the average DVH computed for the spinal cord, brainstem, ipsiand contra-lateral parotids and for the healthy tissue. Fig. 4 shows the average DVH for selected organs at risks showing the potential difference between IMRT and RapidArc in particular cases. Tables 1 and 2 report numerical findings from DVH analysis on PTV, on the main organs at risk and for the healthy tissue. Table 3 summarises results for some complementary organs at risk and for small subgroups of patients. Data are presented as averages over the investigated patients, and errors indicated inter-patient variability at standard deviation level 1.
113
Target coverage and dose homogeneity To simplify the reading of the analysis, data in the tables are reported expressing values as percentages of the dose prescribed to each PTV (e.g. for PTV60, 100% corresponds to 60 Gy), while graphs in the figures are shown with a single normalisation to 100% to 66 Gy. In general, all the techniques resulted in a similar target coverage. RA2 plans achieved the best homogeneity (D5% D95%), while RA1 resulted slightly inferior to both IMRT and RA2. The same trend was observed for D2%, the maximum significant dose, which was inferior to the planning objective of 107%. None of the techniques reached in average the objective on minimum dose, and the same trend was observed as before, with IMRT falling between RA2 and RA1. RapidArc and IMRT showed to be equivalent in terms of conformity index CI95%. This parameter resulted in 1.7 ± 0.2 irrespective of the technique. Spinal cord All plans respected the planning objective of 46 Gy as maximum dose to the spinal cord; RA2 allowed the largest sparing of spinal cord in terms of D2%. Brain stem As for the spinal cord, IMRT, RA1 and RA2 plans showed D2% inferior to the planning objective of 50 Gy, and statistical significance was observed between each technique. With RapidArc, additional reduction of D2% was observed compared to IMRT of 11.7 Gy for RA1 and 13.4 for RA2. Similarly, RapidArc plans showed a further reduction of the volume of the brain stem irradiated at various dose levels, e.g., V20Gy is reduced of 10.4 Gy for RA1 and 12.4 Gy for RA2. Parotids The analysis was carried out for ipsi- and contra-lateral parotids separately and, as expected, larger sparing was observed for the con-
Fig. 1. Dose distributions on axial, coronal and sagittal views for one representative case.
114
Volumetric arc modulation for head and neck radiation therapy
Fig. 2. Mean DVHs for the three PTVs for the global cohort of patients.
Fig. 3. Mean DVHs of the spinal cord, brain stem, parotids and healthy tissue for the global cohort of patients.
tra-lateral glands. Results reported refer to the entire glands, regardless of the (eventual) degree of overlap with the various target volumes. The planning objectives on mean or median dose were in general hard to achieve (met only for the median dose D50% of the contra-lateral glands by RA1 and RA2 plans). In general, as in the case
of the dose delivered to 1/3 or 2/3 of the gland volumes, RA1 and RA2 allowed to reduce the parotids involvement compared to IMRT with a more pronounced efficacy of RA2 compared to RA1. On median dose, (D50%) RA2 allowed an additional sparing of 8.4 Gy and 6 Gy over IMRT in the contra- or ipsi-lateral glands.
E. Vanetti et al. / Radiotherapy and Oncology 92 (2009) 111–117
115
Fig. 4. Mean dose volume histograms for selected complementary organs at risk.
The analysis was repeated for the fraction of parotid glands not included in the target volumes, i.e. for the most superficial portion of the parotids, and in this case much stronger sparing was obviously possible. For the contra-lateral parotids, in average, 90% of the gland volume was outside the PTV, while for the ipsi-lateral case the ratio was reduced to 80%. For the entire patient’s cohort, median doses for IMRT, RA1 and RA2 were 36.7 ± 11.8, 26.3 ± 7.9, and 23.0 ± 8.5 in the ipsi-lateral case and 28.0 ± 9.9, 27.2 ± 4.8, and 21.3 ± 3.2 in the contra-lateral case, respectively.
hypo-pharynx, one larynx and six oro-pharynx). In these limited subgroups of patients, these special organs were either not included or only partially included in the target volumes. From the graphs and the table, it is clear how RapidArc allowed, compared to IMRT, some reduction of mean dose. In particular, RA plans showed a reduction of V50% for the contra-lateral cochlea, the constrictors and the vocal apparatus.
Special cases
The planning objectives for healthy tissue were not formalised in numerical terms but the strategy was to minimise its involvement. In this respect, RA and IMRT presented similar shapes in the DVH of the healthy tissue. No statistically significant difference was observed between the two RapidArc groups. The integral dose was found to be improved with RA in comparison to IMRT with an average reduction of 7% (RA1 or RA2 vs. IMRT).
Fig. 4 reports average dose volume histograms of the base of the tongue (six patients, two from each subgroup), the vocal apparatus (three patients, one hypo-pharynx and two oro-pharynx), the contra-lateral cochlea (six patients, one hypo-pharynx and five oro-pharynx), the constrictors (six patients, two hypo-pharynx, three oropharynx and two larynx) and the mandible (eight patients, one
Healthy tissue
116
Volumetric arc modulation for head and neck radiation therapy
Table 1 Summary of the dosimetric results for the three PTVs. Parameter
Objective [%]
IMRT
RA1
RA2
P
PTV66 Mean [%] D2% [%] D98% [%] D5% D95% [%]
100 107 95 –
100.0 ± 0.0 105.4 ± 1.1 92.4 ± 1.2 9.9 ± 1.6
100.0 ± 0.0 106.2 ± 1.4 91.7 ± 1.6 11.1 ± 2.4
100.0 ± 0.0 104.9 ± 1.2 93.2 ± 1.5 8.8 ± 2.0
n/a a,b,c a,b,c a,b,c
PTV60 Mean [%] D2% [%] D98% [%] D5% D95% [%]
100 107 95 –
100.3 ± 1.6 105.9 ± 2.0 92.3 ± 2.3 10.5 ± 1.8
100.8 ± 1.3 106.4 ± 1.6 92.5 ± 2.3 10.6 ± 2.0
100.7 ± 1.2 105.4 ± 1.6 93.6 ± 2.3 8.7 ± 1.9
– C b,c b,c
PTV54 Mean [%] D2% [%] D98% [%] D5% D95% [%]
100 107 95 –
100.0 ± 2.2 106.2 ± 3.4 92.1 ± 2.8 10.5 ± 2.2
100.0 ± 1.3 106.4 ± 1.7 91.8 ± 2.7 10.6 ± 2.3
99.5 ± 1.6 105.2 ± 1.5 92.1 ± 2.7 9.0 ± 2.1
C C – b,c
Statistical significance (p < 0.05) is reported between couples from paired t-test analysis; a: IMRT vs RA1, b: IMRT vs RA2, c: RA1 vs RA2.
Monitor units and delivery time The number of MU per fraction of 2.2 Gy resulted to be MU/ frIMRT = 1126 ± 333, MU/frRA1 = 463 ± 80 (41% of MU for IMRT), and MU/frRA2 = 584 ± 89 (52% of MU for IMRT). For RapidArc, all individual arcs could be delivered between 70 and 90 s of beamon time. IMRT plans showed values of MUs at least roughly doubled compared to RapidArc and given the multiple field arrangement and the presence of split fields, the delivery time for IMRT is significantly higher since it includes dead times such as the time needed to reposition the gantry and to re-program the linac at every field, giving an overall time for IMRT of 15 min. Discussion This study reports on a comparison of the volumetric modulated arc therapy, RapidArc technique, with single or double coplanar arcs against fixed beam IMRT for head and neck cancer patients. Similar investigations have been published in the recent past. Palma et al. investigated RapidArc progenitor on prostate showing that variable dose rate volumetric arc modulation is beneficial compared to IMRT or constant dose rate [19]. Cozzi et al. [17] and Fogliata et al. [18]. appraised the behaviour of RapidArc on cervix uteri cancer and on small benign brain tumours. In those studies, RapidArc proved to be at least equivalent to IMRT in terms of target coverage while showed benefit
Table 3 Summary of the dosimetric results for special organs at risk not included in the optimisation process. Organ
Patients
Parameter
IMRT
RA1
RA2
Vocal apparatus Base of tongue Mandible
3/29
Contralateral cochlea Constrictors
6/29
Mean [Gy] V50Gy [%] Mean [Gy] V50Gy [%] Mean [Gy] D2% [Gy] Mean [Gy] V50Gy [%] Mean [Gy] V50Gy [%]
59.8 ± 4.8 92.6 ± 12.1 44.1 ± 21.1 49.5 ± 40.4 35.9 ± 6.9 63.8 ± 5.5 22.8 ± 27.6 30.5 ± 47.3 51.8 ± 13.2 75.7 ± 25.6
54.5 ± 6.6 75.7 ± 27.2 38.8 ± 20.9 36.5 ± 39.1 36.9 ± 6.1 62.9 ± 6.3 20.9 ± 22.6 16.1 ± 30.0 48.2 ± 14.1 66.5 ± 28.0
54.8 ± 8.3 73.0 ± 33.4 39.5 ± 21.1 41.3 ± 38.4 33.8 ± 7.4 63.2 ± 6.7 14.2 ± 16.8 5.9 ± 13.1 48.1 ± 13.9 66.3 ± 27.8
6/29 8/29
8/29
in organs at risk sparing. The planning case selected for this investigation was the head and neck. Since it is a demanding indication, it proved to be an ideal indication for IMRT and different strategies have been applied to improve OARs sparing. Compared to IMRT, only RapidArc with two arcs, RA2, allowed a slight improvement in target dose homogeneity and coverage. Both RA1 and RA2 resulted in a systematic reduction of irradiation of spinal cord, brain stem and parotids with statistically significant differences in most of the cases. In addition, single or double RapidArc allowed on average an additional sparing of those organs at risk that are relevant for quality of life and for important acute and late toxicity [15,16] (as base of tongue, vocal apparatus, constrictors and cochlea) but are normally not subject to optimisation objectives even in the absence of explicit constraints in the optimisation. Relevance of the findings of the present study should be validated through a proper clinical protocol that measures toxicity endpoints and control rates (also including the uncertainties derived from fractionated delivery) but this is obviously beyond the scope of a comparative planning study. Some investigations have been performed with experimental dosimetric measurements to assess reliability of RapidArc delivery and its reproducibility, and the first results indicate that RapidArc does not differ from IMRT in this respect [27,28]. Major limitations of, and strategies to minimise biases in, comparative studies, such as the present one, have been extensively discussed elsewhere (e.g. [17]) and shall be similarly applied in the present case. The planning rules were applied as similarly as possible between techniques, dose calculation algorithms and evaluation tools were unified but, even if a lot of care is taken in minimising arbitrary elements, it is impossible to completely control all potential sources of bias and their influence on plan results and comparisons (e.g. the optimisation performed, although using the same objectives, by different planners in different institutes).
Table 2 Summary of the dosimetric results for spinal cord, brain stem, parotids and healthy tissue. Organ
Parameter
Objectives [Gy]
IMRT
RA1
RA2
P
Spinal cord
Mean [Gy] D2% [Gy] Mean [Gy] D2% [Gy] Mean [Gy] D50% [Gy] D33% [Gy] D66% [Gy] Mean [Gy] D50% [Gy] D33% [Gy] D66% [Gy] Mean [Gy] V10Gy [%] DoseIntegral [104 Gy cm3]
– 46 – 50 <26 <30 – – <26 <30 – – – – –
30.8 ± 3.4 42.8 ± 2.1 13.1 ± 10.4 38.2 ± 15.3 40.1 ± 11.6 40.4 ± 13.8 51.0 ± 12.0 30.1 ± 15.6 32.6 ± 8.4 30.1 ± 10.4 41.9 ± 10.5 21.2 ± 10.8 12.2 ± 2.9 33.1 ± 8.4 9.4 ± 3.4
28.2 ± 3.7 43.7 ± 4.1 10.4 ± 8.4 26.5 ± 16.9 36.2 ± 10.8 34.8 ± 14.3 46.4 ± 13.3 25.0 ± 13.8 30.9 ± 7.7 28.4 ± 9.1 38.6 ± 12.1 20.1 ± 6.3 11.5 ± 2.4 30.9 ± 8.0 8.7 ± 2.2
25.3 ± 3.1 39.0 ± 2.6 9.9 ± 8.6 24.8 ± 16.3 34.4 ± 11.1 32.0 ± 15.2 44.7 ± 14.5 22.2 ± 14.6 28.2 ± 6.8 24.1 ± 7.5 36.0 ± 12.1 15.5 ± 4.0 11.4 ± 2.3 31.0 ± 8.0 8.7 ± 2.2
a,b,c b,c a,b a,b,c a,b,c a,b,c a,b,c a,b,c b,c a,b,c a,b,c b,c a,b a,b a,b
Brain stem Ipsi-lateral parotid
Contra-lateral parotid
Healthy tissue
Statistical significance (p < 0.05) is reported between couples from paired t-test analysis; a: IMRT vs RA1, b: IMRT vs RA2, c: RA1 vs RA2.
E. Vanetti et al. / Radiotherapy and Oncology 92 (2009) 111–117
These limitations are, in general, the common features of comparative planning studies, particularly when different optimisation engines or algorithms are considered. It should also be mentioned that, for both techniques, no effort was made to push the optimisation to the maximum achievable but rather the systems were only pushed to reach the planning objectives when possible. Any further improvement shown in the data was obtained ‘‘per se’’ by the algorithms without external driving forces. It is possible that both RA and IMRT might lead to additional sparing of organs with tighter objectives but this was not the scope of the present investigation. It is debatable whether single or multiple arcs should be applied to realise proper volumetric modulated arc techniques. Pioneers applied several arcs to generate modulation patterns with multiple intensity levels [5–8]; alternative commercial solutions to RapidArc will allow multiple arcs and non-coplanar arrangement of arcs. A definitive appraisal of this issue cannot be exploited with single studies since the need for ‘’complex’’ modulation patterns depends on several factors, mostly linked to the clinical indication. Results showed that multiple arcs can improve both the sparing of organs at risk and target coverage. Nevertheless, it is not per se evident if there is a general clinical relevance of further OAR sparing between single and double arcs in terms of expected toxicity or if, as it is likely the case, multiple arcs should be applied as a solution for given critical cases but not as a general standard as single arcs may be largely sufficient in most of the cases. One of the objectives of RapidArc was the capability to deliver treatments in short times. For the cases under investigation, beam-on time was estimated to be less than 1.5 min per arc and considerably less than IMRT delivery times, which are mostly dominated by the large (seven to nine) number of fixed gantry fields. Literature results from other IMAT approaches [7,8,29] show delivery time ranging from 6.3–14 min for C-arm linac based techniques to 11 min for helical tomotherapy depending on the clinical indication. Reduced beam-on and effective treatment times may have a strong impact on the clinical throughput, on the individual management of patients and on the ability to perform systematic image guidance on large groups of patients. Conclusions RapidArc was investigated for head and neck cancer, and RA2 led to significant sparing of organs at risk and healthy tissue sparing with uncompromised target coverage compared to conventional IMRT and RA1. The potential benefit of the better physical dose distribution is combined with the shorter delivery time and smaller number of MUs required. These facts might have an impact on both biology and logistic issues, but whether these reductions translate into clinical benefits will be answered by the clinical outcome studies. Disclosure Dr. L. Cozzi acts as a Scientific Advisor to Varian Medical Systems, and is the Head of Research and Technological Development to Oncology Institute of Southern Switzerland, IOSI, Bellinzona. Acknowledgements The Phase II randomised trial mentioned in the introduction is part of a Research Cooperation Agreement between TMC and Varian Medical Systems. The current investigation was also partially covered by a Varian research grant to IOSI. References [1] Cameron C. Sweeping-window arc therapy: an implementation of rotational IMRT with automatic beam-weight calculation. Phys Med Biol 2005;50:4317–36.
117
[2] Cotrutz C, Kappas C, Webb S. Intensity modulated arc therapy (IMAT) with centrally blocked rotational fields. Phys Med Biol 2000;45: 2185–206. [3] Earl M, Shepard D, Naqvi S, Li X, Zu CX. Inverse planning for intensity modulated arc therapy using direct aperture optimization. Phys Med Biol 2003;21:1075–89. [4] MacKenzie MA, Robinson DM. Intensity modulated arc deliveries approximated by a large number of fixed gantry position sliding window dynamic multileaf collimator fields. Med Phys 2002;29:2359–65. [5] Yu CX. Intensity-modulated arc therapy with dynamic multileaf collimation: an alternative to tomotherapy. Phys Med Biol 1995;40:1435–49. [6] Yu CX, Li XA, Ma L, et al. Clinical implementation of intensity-modulated arc therapy. Int J Radiat Oncol Biol Phys 2002;53:453–63. [7] Duthoy W, De Gersem W, Vergote K, Botenberf T, Derie C, Smeets P, et al. Clinical implementation of intensity modulated arc therapy (IMAT) for rectal cancer. Int J Radiat Oncol Biol Phys 2004;60:794–806. [8] Duthoy W, De Gersem W, Vergote K, Coghe M, Botenberf T, De Deene Y, et al. Whole abdominal radiotherpay (WAPRT) using intensity modulated arc therapy (IMAT): first clinical experience. Int J Radiat Oncol Biol Phys 2003;57:1019–32. [9] Otto K. Volumetric modulated arc therapy: IMRT in a single arc. Med Phys 2008;35:310–7. [10] Gregoire V, De neve W, Eisbruch A, Lee N, Van den Weyngaert D, Van Gestel D. Intensity modulated radiation therapy for head and neck carcinoma. The Oncologist 2007;12:555–64. [11] Lee N, Puri D, Blanco AI, Chao KS. Intensity modulated radiation therapy in head and neck cancers: an update. Head Neck 2007;29:387–400. [12] Popovtzer A, Eisbruch A. Advances in radiation therapy of head and neck cancer. Expert Rev Anticancer Ther 2008;8:633–44. [13] Li Y, Ten Haken R, Eisbruch A. The impact of dose on parotid recovery in head and neck cancer patients treated with radiation therapy. Int J Radiat Oncol Biol Phys 2007;67:660–9. [14] Murdoch-Kinch C, Kim H, Vineberg K, Ship J, Eisbruch A. Dose effect relationships for the submandibular salivary glands and implications for their sparing by intensity modulated radiotherapy. Int J Radiat Oncol Biol Phys 2008;72:373–82. [15] Feng F, Kim H, Lyden T, Feng M, Chepeha D, Eisbruch A. Intensity modulated radiotherapy of head and neck cancer aiming to reduce dysphagia: early doseeffect relationships for the swallowing structures. Int J Radiat Oncol Biol Phys 2007;68:1289–98. [16] Bhandare N, Antonelli P, Morris C, Malayapa R, Mendenhall W. Ototoxicity after radiotherapy for head and neck tumors. Int J Radiat Oncol Biol Phys 2007;67:469–79. [17] Cozzi L, Dinshaw KA, Shrivastava SK, Mahantshetty U, Engineer R, Deshpande DD, et al. A treatment planning study comparing volumetric arc modulation with RapidArc and fixed field IMRT for cervix uteri radiotherapy. Radiother Oncol 2008;89:180–91. [18] Fogliata A, Clivio A, Nicolini G, Vanetti E, Cozzi L. Intensity modulation with photons for benign intracranial tumours. A planning comparison of volumetric single arc, helical arc and fixed gantry techniques. Radiother Oncol 2008 [epub]. [19] Palma D, Vollans E, James K, Nakano S, Moiseenko V, Shaffer R, et al. Volumetric modulated arc therapy for delivery of prostate radiotherapy. Comparison with intensity modulated radiotherapy and threedimensional conformal radiotherapy. Int J Radiat Oncol Biol Phys 2008;72:996–1001. [20] Bragg C, Windate K, Conway J. Clinical implications of the anisotropic analytical algorithm for IMRT treatment planning and verification. Radiother Oncol 2008;86:276–84. [21] Fogliata A, Vanetti E, Albers D, et al. On the dosimetric behaviour of photon dose calculation algorithms in the presence of simple geometric heterogeneities: comparison with Monte Carlo calculations. Phys Med Biol 2007;52:1363–85. [22] Knöös T, Wieslander E, Cozzi L, et al. Comparison of dose calculation algorithms for treatment planning in external photon beam therapy for clinical situations. Phys Med Biol 2006;51:5785–807. [23] Ulmer W, Pyyry J, Kaissl W. A 3D photon superposition/convolution algorithm and its foundation on results of Monte Carlo calculations. Phys Med Biol 2005;50:1767–90. [24] Chui C, LoSasso T, Spirou S. Dose calculation for photon beams with intensity modulation generated by dynamic jaw or multileaf collimations. Med Phys 1994;21:1237–44. [25] Convery D, Rosenbloom M. The generation of intensity-modulated fields for conformal radiotherapy by dynamic collimation. Phys Med Biol 1992;37:1359–74. [26] Spirou S, Chui C. A gradient inverse planning algorithm with dose-volume constrains. Med Phys 1998;25:321–33. [27] S. Korreman, J. Medin, F. Kjaer-Kreistoffersen. Dosimetric verification of RapidArc treatment delivery. Acta Oncol 2008 [epub]. [28] Nicolini G, Vanetti E, Clivio A, Fogliata A, Korreman S, Bocanek J, et al. The GLAaS algorithm for portal dosimetry and quality assurance of RapidArc, an intensity modulated rotational therapy. Radiat Oncol 2008;3:24. [29] Bijdekerke P, Verellen D, Tournel K, Vinh-Hung V, Somers F, Biesman P, et al. Tomotherapy: implications on daily workload and scheduling patients. Radiother Oncol 2008;86:224–30.
Contents lists available at ScienceDirect
Radiotherapy and Oncology journal homepage: www.thegreenjournal.com
Head and neck IMRT
Volumetric modulated arc radiotherapy for carcinomas of the oro-pharynx, hypo-pharynx and larynx: A treatment planning comparison with fixed field IMRT Eugenio Vanetti a, Alessandro Clivio a, Giorgia Nicolini a, Antonella Fogliata a, Sarbani Ghosh-Laskar b, Jai Prakash Agarwal b, Ritu Raj Upreti b, Ashwini Budrukkar b, Vedang Murthy b, Deepak Dattatray Deshpande b, Shyam Kishore Shrivastava b, Ketayun Ardeshir Dinshaw b, Luca Cozzi a,* a b
Oncology Institute of Southern Switzerland, Radiation Oncology Department, Bellinzona, Switzerland Departments of Radiation Oncology & Medical Physics, Tata Memorial Hospital, Mumbai, India
a r t i c l e
i n f o
Article history: Received 2 September 2008 Received in revised form 23 December 2008 Accepted 26 December 2008 Available online 20 January 2009 Keywords: RapidArc IMRT Head and neck radiation therapy
a b s t r a c t Purpose: A planning study was performed to evaluate the performance of volumetric modulated arc radiotherapy on head and neck cancer patients. Conventional fixed field IMRT was used as a benchmark. Methods and materials: CT datasets of 29 patients with squamous cell carcinoma of the oro-pharynx, hypo-pharynx and larynx were included. Plans for fixed beam IMRT, single (RA1) and double (RA2) modulated arcs with the RapidArc technique were optimised. Dose prescription was set to 66 Gy to the primary tumour (at 2.2 Gy/fraction), 60 Gy to intermediate-risk nodes and 54 Gy to low-risk nodal levels. The planning objectives for PTV were minimum dose >95%, and maximum dose <107%. Maximum dose to spinal cord was limited to 46 Gy, maximum to brain stem to 50 Gy. For parotids, mean dose <26 Gy (or median <30 Gy) was assumed as the objective. The MU and delivery time were scored to measure expected treatment efficiency. Results: Target coverage and homogeneity results improved with RA2 plans compared to both RA1 and IMRT. All the techniques fulfilled the objectives on maximum dose, while small deviations were observed on minimum dose for PTV. The conformity index (CI95%) was 1.7 ± 0.2 for all the three techniques. RA2 allowed a reduction of D2% to spinal cord of 3 Gy compared to IMRT (RA1 D2% increased it of 1 Gy). On brain stem, D2% was reduced from 12 Gy (RA1 vs. IMRT) to 13.5 Gy (RA2 vs. IMRT). The mean dose to ipsi-lateral parotids was reduced from 40 Gy (IMRT) to 36.2 Gy (RA1) and 34.4 Gy (RA2). The mean dose to the contra-lateral gland ranged from 32.6 Gy (IMRT) to 30.9 Gy (RA1) and 28.2 Gy (RA2). Conclusion: RapidArc was investigated for head and neck cancer. RA1 and RA2 showed some improvements in organs at risk and healthy tissue sparing, while only RA2 offered improved target coverage with respect to conventional IMRT. Ó 2009 Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology 92 (2009) 111–117
The aim of the present study was to investigate the potential clinical role for head and neck cancer patients of RapidArc, the novel radiation treatment technique (Varian Medical Systems), which is based on volumetric intensity-modulated arc delivery, as opposed to intensity modulation which uses fixed gantry beams. RapidArc falls into the category of intensity modulation therapy with arcs (IMAT) [1–4]. The Yu’s group established the benefit of using multiple modulated arcs for complex cases [5,6]. The Ghent group applied IMAT techniques with multiple non-coplanar beams to pelvic treatments [7,8] proving equivalent or superior target coverage and improved sparing of OARs compared to conventional conformal treatments.
* Corresponding author. Address: Oncology Institute of Southern Switzerland, Radiation Oncology Department, Medical Physics Unit, 6504 Bellinzona, Switzerland. E-mail address: [email protected] (L. Cozzi). 0167-8140/$ - see front matter Ó 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.radonc.2008.12.008
RapidArc is a technique based on an investigation from K. Otto [9] and it aims to (i) improve OARs and healthy tissue sparing compared to other solutions; (ii) maintain or improve the same degree of target coverage; and (iii) reduce beam-on time per fraction. Faster treatments could have a clinical impact on patients in terms of comfort on couch, immobility and minimisation of internal organ displacement. It could also allow more time for imaging procedures allowing, in perspective, routine application of adaptive treatment strategies when changes are observed due to response to radiation as in patients with head and neck cancer. IMRT in head and neck cancer patients has been largely investigated at both planning and clinical levels. Excellent reviews for treatment outcome and major toxicity patterns can be found in Gregoire et al. [10], Lee et al. [11] and Popovtzer et al. [12]. On the toxicity side, besides the major attention given to spinal cord and brain stem (with toxicity thresholds set in the proximity of 45–50 Gy for the first and at 50 Gy for the second), it is generally
112
Volumetric arc modulation for head and neck radiation therapy
known that, for parotids, mean doses inferior to 25–30 Gy correlate well with substantial recovery of function within two years [13] (higher thresholds were observed for sub-mandibular glands in the range of 39 Gy [14]). To reduce dysphagia [15], sparing of constrictors (with mean dose below 60 Gy) highly correlated with improved swallowing, laryngeal elevation and epiglottic inversion. To manage ototoxicity [16], the mean dose to cochlea proved to be a highly significant factor with some threshold effect in the order of 60 Gy between high and low risks of auditory defects. Volumetric IMAT has already been investigated for prostate, small brain tumours and cervix uteri cancer [17–19], i.e. on relatively simple clinical cases; head and neck is an ideal advanced benchmark for assessment of its conformal avoidance capabilities since the anatomical features of this location require highly sophisticated techniques to ensure adequate treatments. The present study was initiated as a side investigation in the framework of a larger Phase II trial activated at Tata Memorial Centre (TMC) in Mumbai to investigate the role of IMRT vs 3D conformal radiotherapy in squamous cell carcinoma (SCC) patients at AJCC stages T1-3, N0-2b, and M0.
Material and methods Patient selection and planning objectives CT data (3 mm slice thickness) for a group of twenty-nine consecutively treated patients from the TMC protocol were selected for the purposes of planning. These patients had been randomized and treated on the IMRT arm of the protocol. The RapidArc planning was carried out on the same dataset at the Oncology Institute of Southern Switzerland. Fourteen tumours were localised in the oro-pharynx district, eight in the hypo-pharynx and seven in the larynx. Fourteen patients presented T3 stage, twelve T2 stage and two T1 stage. Nodal involvement was mostly N0 (16), N1 (6) and N2 (7) stages; all patients presented no distant metastasis (M0). The main organs at risk (OAR) considered for all patients were ipsi- and contra-lateral parotids, spinal cord and brain stem. For some of the patients, additional organs were outlined by radiation oncologists depending on the indication including oral cavity, oesophageal constrictors, base of the tongue, cochlea, mandible, and vocal apparatus. These additional OARs were defined only for those patients where potential sparing was possible due to the target conformation. The healthy tissue was defined as the patient’s volume covered by the CT scan minus the envelope of the various target volumes (PTV). Various PTVs were defined from the respective clinical target volumes (CTVs) adding 7 mm margins with 3D expansion. Dose prescription was set to 66 Gy at 2.2 Gy/fraction to the PTV including the primary tumour and the lymph-nodal metastases (PTV66 – high risk). Two additional elective PTVs were defined to be irradiated at 60 Gy (intermediate risk) and 54 Gy (low risk, sub-clinical disease) in the nodal regions (PTV60 and PTV54). All volumes were to be simultaneously treated according to the simultaneous integrated boost (SIB) approach. Due to target definition criteria, PTVs did not overlap and were not mutually included. All plans were normalised to the mean dose of PTV66. The mean volume of target volumes was PTV66: 342 ± 182 cm3, PTV60: 107 ± 81 cm3, and PTV54: 95 ± 54 cm3 for the global group. For all the PTVs, plans aimed to achieve minimum dose larger than 95% of the prescribed dose and a maximum lower than 107%. For the spinal cord, a maximum dose of 46 Gy was allowed. For the brain stem, the limit was set to 50 Gy. In the case of parotids, the planning objectives aimed to keep the mean dose below 26 Gy (or
D50% < 30 Gy). For other organs at risk, the planning strategy was to minimise their involvement but no specific constraint was set. Planning techniques Three sets of plans were compared in this study, all designed on the Varian Eclipse treatment planning system (TPS) with 6 MV photon beams from a Varian Clinac equipped with a Millennium Multileaf Collimator (MLC) with 120 leaves (spatial resolution of 5 mm at isocentre for the central 20 cm and of 10 mm in the outer 2 10 cm, maximum leaf speed of 2.5 cm/s and leaf transmission of 1.8%). Plans for RapidArc were optimised selecting a maximum DR of 600 MU/ min, and a fixed DR of 300 MU/min was selected for IMRT. The Anisotropic Analytical Algorithm (AAA, version 8.6.02) photon dose calculation algorithm was used for all cases [20–23]. The dose calculation grid was set to 2.5 mm. RapidArc optimisation was performed with version 8.6.05. Details for each planning method are as follows IMRT Reference plans were computed by selecting the ‘conventional’ intensity modulation approach as a benchmark, with fixed gantry and intensity-modulated beams delivering the dose by means of the sliding window approach [24–26]. Plans were individually optimised using seven or nine coplanar fields. The modulation fluences used in the present analysis are equivalent to the fluences approved for patient treatment, while final dose calculation was performed using the AAA, including heterogeneity management, instead of the Pencil Beam (used at TMC) for coherence with the RapidArc condition. MUs were kept fixed at the value from original calculation. It should be noted that in Eclipse, the optimisation process to generate the fluences and the final dose calculation are completely disentangled and therefore the determination of the optimal fluence is in no way influenced by the algorithm adopted afterwards. RapidArc (RA) To achieve the desired level of modulation required, the instantaneous dose rate (DR), MLC leaf positions and the gantry rotational speed were continuously varied by the RapidArc optimiser. To minimise the contribution of tongue and groove effect, the collimator rotation in RapidArc was kept fixed to a value different from zero. In the present study, the collimator was rotated to 40°. Details of the RapidArc process can be found in [9,15]. Two sets of plans were optimised and analysed. RA1 consisting of a single 360° rotation and RA2 consisting of two coplanar arcs of 360° were optimised simultaneously, to be delivered with opposite rotation (clock- and counter clock-wise). The application of two coplanar arcs aims to increase the modulation factor during optimisation. In fact, since each individual arc is limited to a sequence of 177 control points (i.e. ‘’elementary’’ fields), the application of two independent arcs, simultaneously optimised, could allow the optimiser to achieve higher target homogeneity and lower OARs involvement at the same time, as seen in other IMAT applications [5–8]. In the previous investigations on RapidArc [17,18], the relative simplicity of cases did not require investigation of this feature of the optimiser. Evaluation tools Quantitative evaluation of plans was performed by means of standard Dose-Volume Histogram (DVH). For PTV, the values of D98% and D2% (dose received by the 98%, and 2% of the volume) were defined as metrics for minimum and maximum doses in association to V95% V107% (the volume receiving at least 95% or at most
E. Vanetti et al. / Radiotherapy and Oncology 92 (2009) 111–117
107% of the prescribed dose). The target homogeneity was expressed by D5% D95% (difference between the dose covering 5% and 95% of the PTV). The degree of conformality of the plans was measured with a Conformity Index, CI95%, which is defined as the ratio between the patient volume receiving at least 95% of the prescribed dose and the volume of the PTV. For OARs, the analysis included the mean dose, the maximum dose expressed as D2% and a set of appropriate VX and DY values. For Healthy Tissue, the integral dose, ‘‘DoseInt”, is defined as the integral of the absorbed dose extended to over all voxels excluding those within the target volume (DoseInt dimensions are Gy cm3). This was reported together with the observed mean dose and some representative VxGy value. The average cumulative DVH for PTV, OARs and healthy tissue were built from the individual DVHs obtained by averaging the corresponding volumes at each dose bin (0.01 Gy in this case). To appraise the difference between the techniques, the paired, twotailed Student’s t-test was applied. Data were considered statistically significant for p < 0.05. Results Dose distributions are shown for one example in Fig. 1 for axial sagittal and coronal views. Fig. 2 shows the average DVH for all the PTVs comparing the three techniques for the entire patients cohort. Fig. 3 reports the average DVH computed for the spinal cord, brainstem, ipsiand contra-lateral parotids and for the healthy tissue. Fig. 4 shows the average DVH for selected organs at risks showing the potential difference between IMRT and RapidArc in particular cases. Tables 1 and 2 report numerical findings from DVH analysis on PTV, on the main organs at risk and for the healthy tissue. Table 3 summarises results for some complementary organs at risk and for small subgroups of patients. Data are presented as averages over the investigated patients, and errors indicated inter-patient variability at standard deviation level 1.
113
Target coverage and dose homogeneity To simplify the reading of the analysis, data in the tables are reported expressing values as percentages of the dose prescribed to each PTV (e.g. for PTV60, 100% corresponds to 60 Gy), while graphs in the figures are shown with a single normalisation to 100% to 66 Gy. In general, all the techniques resulted in a similar target coverage. RA2 plans achieved the best homogeneity (D5% D95%), while RA1 resulted slightly inferior to both IMRT and RA2. The same trend was observed for D2%, the maximum significant dose, which was inferior to the planning objective of 107%. None of the techniques reached in average the objective on minimum dose, and the same trend was observed as before, with IMRT falling between RA2 and RA1. RapidArc and IMRT showed to be equivalent in terms of conformity index CI95%. This parameter resulted in 1.7 ± 0.2 irrespective of the technique. Spinal cord All plans respected the planning objective of 46 Gy as maximum dose to the spinal cord; RA2 allowed the largest sparing of spinal cord in terms of D2%. Brain stem As for the spinal cord, IMRT, RA1 and RA2 plans showed D2% inferior to the planning objective of 50 Gy, and statistical significance was observed between each technique. With RapidArc, additional reduction of D2% was observed compared to IMRT of 11.7 Gy for RA1 and 13.4 for RA2. Similarly, RapidArc plans showed a further reduction of the volume of the brain stem irradiated at various dose levels, e.g., V20Gy is reduced of 10.4 Gy for RA1 and 12.4 Gy for RA2. Parotids The analysis was carried out for ipsi- and contra-lateral parotids separately and, as expected, larger sparing was observed for the con-
Fig. 1. Dose distributions on axial, coronal and sagittal views for one representative case.
114
Volumetric arc modulation for head and neck radiation therapy
Fig. 2. Mean DVHs for the three PTVs for the global cohort of patients.
Fig. 3. Mean DVHs of the spinal cord, brain stem, parotids and healthy tissue for the global cohort of patients.
tra-lateral glands. Results reported refer to the entire glands, regardless of the (eventual) degree of overlap with the various target volumes. The planning objectives on mean or median dose were in general hard to achieve (met only for the median dose D50% of the contra-lateral glands by RA1 and RA2 plans). In general, as in the case
of the dose delivered to 1/3 or 2/3 of the gland volumes, RA1 and RA2 allowed to reduce the parotids involvement compared to IMRT with a more pronounced efficacy of RA2 compared to RA1. On median dose, (D50%) RA2 allowed an additional sparing of 8.4 Gy and 6 Gy over IMRT in the contra- or ipsi-lateral glands.
E. Vanetti et al. / Radiotherapy and Oncology 92 (2009) 111–117
115
Fig. 4. Mean dose volume histograms for selected complementary organs at risk.
The analysis was repeated for the fraction of parotid glands not included in the target volumes, i.e. for the most superficial portion of the parotids, and in this case much stronger sparing was obviously possible. For the contra-lateral parotids, in average, 90% of the gland volume was outside the PTV, while for the ipsi-lateral case the ratio was reduced to 80%. For the entire patient’s cohort, median doses for IMRT, RA1 and RA2 were 36.7 ± 11.8, 26.3 ± 7.9, and 23.0 ± 8.5 in the ipsi-lateral case and 28.0 ± 9.9, 27.2 ± 4.8, and 21.3 ± 3.2 in the contra-lateral case, respectively.
hypo-pharynx, one larynx and six oro-pharynx). In these limited subgroups of patients, these special organs were either not included or only partially included in the target volumes. From the graphs and the table, it is clear how RapidArc allowed, compared to IMRT, some reduction of mean dose. In particular, RA plans showed a reduction of V50% for the contra-lateral cochlea, the constrictors and the vocal apparatus.
Special cases
The planning objectives for healthy tissue were not formalised in numerical terms but the strategy was to minimise its involvement. In this respect, RA and IMRT presented similar shapes in the DVH of the healthy tissue. No statistically significant difference was observed between the two RapidArc groups. The integral dose was found to be improved with RA in comparison to IMRT with an average reduction of 7% (RA1 or RA2 vs. IMRT).
Fig. 4 reports average dose volume histograms of the base of the tongue (six patients, two from each subgroup), the vocal apparatus (three patients, one hypo-pharynx and two oro-pharynx), the contra-lateral cochlea (six patients, one hypo-pharynx and five oro-pharynx), the constrictors (six patients, two hypo-pharynx, three oropharynx and two larynx) and the mandible (eight patients, one
Healthy tissue
116
Volumetric arc modulation for head and neck radiation therapy
Table 1 Summary of the dosimetric results for the three PTVs. Parameter
Objective [%]
IMRT
RA1
RA2
P
PTV66 Mean [%] D2% [%] D98% [%] D5% D95% [%]
100 107 95 –
100.0 ± 0.0 105.4 ± 1.1 92.4 ± 1.2 9.9 ± 1.6
100.0 ± 0.0 106.2 ± 1.4 91.7 ± 1.6 11.1 ± 2.4
100.0 ± 0.0 104.9 ± 1.2 93.2 ± 1.5 8.8 ± 2.0
n/a a,b,c a,b,c a,b,c
PTV60 Mean [%] D2% [%] D98% [%] D5% D95% [%]
100 107 95 –
100.3 ± 1.6 105.9 ± 2.0 92.3 ± 2.3 10.5 ± 1.8
100.8 ± 1.3 106.4 ± 1.6 92.5 ± 2.3 10.6 ± 2.0
100.7 ± 1.2 105.4 ± 1.6 93.6 ± 2.3 8.7 ± 1.9
– C b,c b,c
PTV54 Mean [%] D2% [%] D98% [%] D5% D95% [%]
100 107 95 –
100.0 ± 2.2 106.2 ± 3.4 92.1 ± 2.8 10.5 ± 2.2
100.0 ± 1.3 106.4 ± 1.7 91.8 ± 2.7 10.6 ± 2.3
99.5 ± 1.6 105.2 ± 1.5 92.1 ± 2.7 9.0 ± 2.1
C C – b,c
Statistical significance (p < 0.05) is reported between couples from paired t-test analysis; a: IMRT vs RA1, b: IMRT vs RA2, c: RA1 vs RA2.
Monitor units and delivery time The number of MU per fraction of 2.2 Gy resulted to be MU/ frIMRT = 1126 ± 333, MU/frRA1 = 463 ± 80 (41% of MU for IMRT), and MU/frRA2 = 584 ± 89 (52% of MU for IMRT). For RapidArc, all individual arcs could be delivered between 70 and 90 s of beamon time. IMRT plans showed values of MUs at least roughly doubled compared to RapidArc and given the multiple field arrangement and the presence of split fields, the delivery time for IMRT is significantly higher since it includes dead times such as the time needed to reposition the gantry and to re-program the linac at every field, giving an overall time for IMRT of 15 min. Discussion This study reports on a comparison of the volumetric modulated arc therapy, RapidArc technique, with single or double coplanar arcs against fixed beam IMRT for head and neck cancer patients. Similar investigations have been published in the recent past. Palma et al. investigated RapidArc progenitor on prostate showing that variable dose rate volumetric arc modulation is beneficial compared to IMRT or constant dose rate [19]. Cozzi et al. [17] and Fogliata et al. [18]. appraised the behaviour of RapidArc on cervix uteri cancer and on small benign brain tumours. In those studies, RapidArc proved to be at least equivalent to IMRT in terms of target coverage while showed benefit
Table 3 Summary of the dosimetric results for special organs at risk not included in the optimisation process. Organ
Patients
Parameter
IMRT
RA1
RA2
Vocal apparatus Base of tongue Mandible
3/29
Contralateral cochlea Constrictors
6/29
Mean [Gy] V50Gy [%] Mean [Gy] V50Gy [%] Mean [Gy] D2% [Gy] Mean [Gy] V50Gy [%] Mean [Gy] V50Gy [%]
59.8 ± 4.8 92.6 ± 12.1 44.1 ± 21.1 49.5 ± 40.4 35.9 ± 6.9 63.8 ± 5.5 22.8 ± 27.6 30.5 ± 47.3 51.8 ± 13.2 75.7 ± 25.6
54.5 ± 6.6 75.7 ± 27.2 38.8 ± 20.9 36.5 ± 39.1 36.9 ± 6.1 62.9 ± 6.3 20.9 ± 22.6 16.1 ± 30.0 48.2 ± 14.1 66.5 ± 28.0
54.8 ± 8.3 73.0 ± 33.4 39.5 ± 21.1 41.3 ± 38.4 33.8 ± 7.4 63.2 ± 6.7 14.2 ± 16.8 5.9 ± 13.1 48.1 ± 13.9 66.3 ± 27.8
6/29 8/29
8/29
in organs at risk sparing. The planning case selected for this investigation was the head and neck. Since it is a demanding indication, it proved to be an ideal indication for IMRT and different strategies have been applied to improve OARs sparing. Compared to IMRT, only RapidArc with two arcs, RA2, allowed a slight improvement in target dose homogeneity and coverage. Both RA1 and RA2 resulted in a systematic reduction of irradiation of spinal cord, brain stem and parotids with statistically significant differences in most of the cases. In addition, single or double RapidArc allowed on average an additional sparing of those organs at risk that are relevant for quality of life and for important acute and late toxicity [15,16] (as base of tongue, vocal apparatus, constrictors and cochlea) but are normally not subject to optimisation objectives even in the absence of explicit constraints in the optimisation. Relevance of the findings of the present study should be validated through a proper clinical protocol that measures toxicity endpoints and control rates (also including the uncertainties derived from fractionated delivery) but this is obviously beyond the scope of a comparative planning study. Some investigations have been performed with experimental dosimetric measurements to assess reliability of RapidArc delivery and its reproducibility, and the first results indicate that RapidArc does not differ from IMRT in this respect [27,28]. Major limitations of, and strategies to minimise biases in, comparative studies, such as the present one, have been extensively discussed elsewhere (e.g. [17]) and shall be similarly applied in the present case. The planning rules were applied as similarly as possible between techniques, dose calculation algorithms and evaluation tools were unified but, even if a lot of care is taken in minimising arbitrary elements, it is impossible to completely control all potential sources of bias and their influence on plan results and comparisons (e.g. the optimisation performed, although using the same objectives, by different planners in different institutes).
Table 2 Summary of the dosimetric results for spinal cord, brain stem, parotids and healthy tissue. Organ
Parameter
Objectives [Gy]
IMRT
RA1
RA2
P
Spinal cord
Mean [Gy] D2% [Gy] Mean [Gy] D2% [Gy] Mean [Gy] D50% [Gy] D33% [Gy] D66% [Gy] Mean [Gy] D50% [Gy] D33% [Gy] D66% [Gy] Mean [Gy] V10Gy [%] DoseIntegral [104 Gy cm3]
– 46 – 50 <26 <30 – – <26 <30 – – – – –
30.8 ± 3.4 42.8 ± 2.1 13.1 ± 10.4 38.2 ± 15.3 40.1 ± 11.6 40.4 ± 13.8 51.0 ± 12.0 30.1 ± 15.6 32.6 ± 8.4 30.1 ± 10.4 41.9 ± 10.5 21.2 ± 10.8 12.2 ± 2.9 33.1 ± 8.4 9.4 ± 3.4
28.2 ± 3.7 43.7 ± 4.1 10.4 ± 8.4 26.5 ± 16.9 36.2 ± 10.8 34.8 ± 14.3 46.4 ± 13.3 25.0 ± 13.8 30.9 ± 7.7 28.4 ± 9.1 38.6 ± 12.1 20.1 ± 6.3 11.5 ± 2.4 30.9 ± 8.0 8.7 ± 2.2
25.3 ± 3.1 39.0 ± 2.6 9.9 ± 8.6 24.8 ± 16.3 34.4 ± 11.1 32.0 ± 15.2 44.7 ± 14.5 22.2 ± 14.6 28.2 ± 6.8 24.1 ± 7.5 36.0 ± 12.1 15.5 ± 4.0 11.4 ± 2.3 31.0 ± 8.0 8.7 ± 2.2
a,b,c b,c a,b a,b,c a,b,c a,b,c a,b,c a,b,c b,c a,b,c a,b,c b,c a,b a,b a,b
Brain stem Ipsi-lateral parotid
Contra-lateral parotid
Healthy tissue
Statistical significance (p < 0.05) is reported between couples from paired t-test analysis; a: IMRT vs RA1, b: IMRT vs RA2, c: RA1 vs RA2.
E. Vanetti et al. / Radiotherapy and Oncology 92 (2009) 111–117
These limitations are, in general, the common features of comparative planning studies, particularly when different optimisation engines or algorithms are considered. It should also be mentioned that, for both techniques, no effort was made to push the optimisation to the maximum achievable but rather the systems were only pushed to reach the planning objectives when possible. Any further improvement shown in the data was obtained ‘‘per se’’ by the algorithms without external driving forces. It is possible that both RA and IMRT might lead to additional sparing of organs with tighter objectives but this was not the scope of the present investigation. It is debatable whether single or multiple arcs should be applied to realise proper volumetric modulated arc techniques. Pioneers applied several arcs to generate modulation patterns with multiple intensity levels [5–8]; alternative commercial solutions to RapidArc will allow multiple arcs and non-coplanar arrangement of arcs. A definitive appraisal of this issue cannot be exploited with single studies since the need for ‘’complex’’ modulation patterns depends on several factors, mostly linked to the clinical indication. Results showed that multiple arcs can improve both the sparing of organs at risk and target coverage. Nevertheless, it is not per se evident if there is a general clinical relevance of further OAR sparing between single and double arcs in terms of expected toxicity or if, as it is likely the case, multiple arcs should be applied as a solution for given critical cases but not as a general standard as single arcs may be largely sufficient in most of the cases. One of the objectives of RapidArc was the capability to deliver treatments in short times. For the cases under investigation, beam-on time was estimated to be less than 1.5 min per arc and considerably less than IMRT delivery times, which are mostly dominated by the large (seven to nine) number of fixed gantry fields. Literature results from other IMAT approaches [7,8,29] show delivery time ranging from 6.3–14 min for C-arm linac based techniques to 11 min for helical tomotherapy depending on the clinical indication. Reduced beam-on and effective treatment times may have a strong impact on the clinical throughput, on the individual management of patients and on the ability to perform systematic image guidance on large groups of patients. Conclusions RapidArc was investigated for head and neck cancer, and RA2 led to significant sparing of organs at risk and healthy tissue sparing with uncompromised target coverage compared to conventional IMRT and RA1. The potential benefit of the better physical dose distribution is combined with the shorter delivery time and smaller number of MUs required. These facts might have an impact on both biology and logistic issues, but whether these reductions translate into clinical benefits will be answered by the clinical outcome studies. Disclosure Dr. L. Cozzi acts as a Scientific Advisor to Varian Medical Systems, and is the Head of Research and Technological Development to Oncology Institute of Southern Switzerland, IOSI, Bellinzona. Acknowledgements The Phase II randomised trial mentioned in the introduction is part of a Research Cooperation Agreement between TMC and Varian Medical Systems. The current investigation was also partially covered by a Varian research grant to IOSI. References [1] Cameron C. Sweeping-window arc therapy: an implementation of rotational IMRT with automatic beam-weight calculation. Phys Med Biol 2005;50:4317–36.
117
[2] Cotrutz C, Kappas C, Webb S. Intensity modulated arc therapy (IMAT) with centrally blocked rotational fields. Phys Med Biol 2000;45: 2185–206. [3] Earl M, Shepard D, Naqvi S, Li X, Zu CX. Inverse planning for intensity modulated arc therapy using direct aperture optimization. Phys Med Biol 2003;21:1075–89. [4] MacKenzie MA, Robinson DM. Intensity modulated arc deliveries approximated by a large number of fixed gantry position sliding window dynamic multileaf collimator fields. Med Phys 2002;29:2359–65. [5] Yu CX. Intensity-modulated arc therapy with dynamic multileaf collimation: an alternative to tomotherapy. Phys Med Biol 1995;40:1435–49. [6] Yu CX, Li XA, Ma L, et al. Clinical implementation of intensity-modulated arc therapy. Int J Radiat Oncol Biol Phys 2002;53:453–63. [7] Duthoy W, De Gersem W, Vergote K, Botenberf T, Derie C, Smeets P, et al. Clinical implementation of intensity modulated arc therapy (IMAT) for rectal cancer. Int J Radiat Oncol Biol Phys 2004;60:794–806. [8] Duthoy W, De Gersem W, Vergote K, Coghe M, Botenberf T, De Deene Y, et al. Whole abdominal radiotherpay (WAPRT) using intensity modulated arc therapy (IMAT): first clinical experience. Int J Radiat Oncol Biol Phys 2003;57:1019–32. [9] Otto K. Volumetric modulated arc therapy: IMRT in a single arc. Med Phys 2008;35:310–7. [10] Gregoire V, De neve W, Eisbruch A, Lee N, Van den Weyngaert D, Van Gestel D. Intensity modulated radiation therapy for head and neck carcinoma. The Oncologist 2007;12:555–64. [11] Lee N, Puri D, Blanco AI, Chao KS. Intensity modulated radiation therapy in head and neck cancers: an update. Head Neck 2007;29:387–400. [12] Popovtzer A, Eisbruch A. Advances in radiation therapy of head and neck cancer. Expert Rev Anticancer Ther 2008;8:633–44. [13] Li Y, Ten Haken R, Eisbruch A. The impact of dose on parotid recovery in head and neck cancer patients treated with radiation therapy. Int J Radiat Oncol Biol Phys 2007;67:660–9. [14] Murdoch-Kinch C, Kim H, Vineberg K, Ship J, Eisbruch A. Dose effect relationships for the submandibular salivary glands and implications for their sparing by intensity modulated radiotherapy. Int J Radiat Oncol Biol Phys 2008;72:373–82. [15] Feng F, Kim H, Lyden T, Feng M, Chepeha D, Eisbruch A. Intensity modulated radiotherapy of head and neck cancer aiming to reduce dysphagia: early doseeffect relationships for the swallowing structures. Int J Radiat Oncol Biol Phys 2007;68:1289–98. [16] Bhandare N, Antonelli P, Morris C, Malayapa R, Mendenhall W. Ototoxicity after radiotherapy for head and neck tumors. Int J Radiat Oncol Biol Phys 2007;67:469–79. [17] Cozzi L, Dinshaw KA, Shrivastava SK, Mahantshetty U, Engineer R, Deshpande DD, et al. A treatment planning study comparing volumetric arc modulation with RapidArc and fixed field IMRT for cervix uteri radiotherapy. Radiother Oncol 2008;89:180–91. [18] Fogliata A, Clivio A, Nicolini G, Vanetti E, Cozzi L. Intensity modulation with photons for benign intracranial tumours. A planning comparison of volumetric single arc, helical arc and fixed gantry techniques. Radiother Oncol 2008 [epub]. [19] Palma D, Vollans E, James K, Nakano S, Moiseenko V, Shaffer R, et al. Volumetric modulated arc therapy for delivery of prostate radiotherapy. Comparison with intensity modulated radiotherapy and threedimensional conformal radiotherapy. Int J Radiat Oncol Biol Phys 2008;72:996–1001. [20] Bragg C, Windate K, Conway J. Clinical implications of the anisotropic analytical algorithm for IMRT treatment planning and verification. Radiother Oncol 2008;86:276–84. [21] Fogliata A, Vanetti E, Albers D, et al. On the dosimetric behaviour of photon dose calculation algorithms in the presence of simple geometric heterogeneities: comparison with Monte Carlo calculations. Phys Med Biol 2007;52:1363–85. [22] Knöös T, Wieslander E, Cozzi L, et al. Comparison of dose calculation algorithms for treatment planning in external photon beam therapy for clinical situations. Phys Med Biol 2006;51:5785–807. [23] Ulmer W, Pyyry J, Kaissl W. A 3D photon superposition/convolution algorithm and its foundation on results of Monte Carlo calculations. Phys Med Biol 2005;50:1767–90. [24] Chui C, LoSasso T, Spirou S. Dose calculation for photon beams with intensity modulation generated by dynamic jaw or multileaf collimations. Med Phys 1994;21:1237–44. [25] Convery D, Rosenbloom M. The generation of intensity-modulated fields for conformal radiotherapy by dynamic collimation. Phys Med Biol 1992;37:1359–74. [26] Spirou S, Chui C. A gradient inverse planning algorithm with dose-volume constrains. Med Phys 1998;25:321–33. [27] S. Korreman, J. Medin, F. Kjaer-Kreistoffersen. Dosimetric verification of RapidArc treatment delivery. Acta Oncol 2008 [epub]. [28] Nicolini G, Vanetti E, Clivio A, Fogliata A, Korreman S, Bocanek J, et al. The GLAaS algorithm for portal dosimetry and quality assurance of RapidArc, an intensity modulated rotational therapy. Radiat Oncol 2008;3:24. [29] Bijdekerke P, Verellen D, Tournel K, Vinh-Hung V, Somers F, Biesman P, et al. Tomotherapy: implications on daily workload and scheduling patients. Radiother Oncol 2008;86:224–30.