- Email: [email protected]
Volumetric-modulated arc radiotherapy for carcinomas of the anal canal: A treatment planning comparison with fixed field IMRT
Radiotherapy and Oncology 92 (2009) 118–124
Contents lists available at ScienceDirect
Radiotherapy and Oncology journal homepage: www.thegreenjournal.com
Anal cancer IMRT
Volumetric-modulated arc radiotherapy for carcinomas of the anal canal: A treatment planning comparison with fixed field IMRT Alessandro Clivio a, Antonella Fogliata a, Alessandra Franzetti-Pellanda b, Giorgia Nicolini a, Eugenio Vanetti a, Rolf Wyttenbach c,d, Luca Cozzi a,* a
Medical Physics, Oncology Institute of Southern Switzerland, Bellinzona, Switzerland Radiation Oncology Department, Oncology Institute of Southern Switzerland, Bellinzona, Switzerland c Ospedale San Giovanni, Radiology, Bellinzona, Switzerland d University of Bern, Bern, Switzerland b
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 30 January 2009 Keywords: RapidArc IMRT Anal canal radiation therapy Volumetric-modulated arc therapy
a b s t r a c t Purpose: A treatment planning study was performed to compare volumetric-modulated arc radiotherapy against conventional fixed field IMRT. Materials and methods: CT datasets of 10 patients affected by carcinoma of the anal canal were included and five plans were generated for each case: fixed beam IMRT, single (RA1)- and double (RA2)-modulated arcs with the RapidArc technique. Dose prescription was set according to a simultaneous integrated boost strategy to 59.4 Gy to the primary tumour PTVI (at 1.8 Gy/fraction) and to 49.5 Gy to risk area including inguinal nodes, PTVII. Planning objectives for PTV were minimum dose >95%, maximum dose < 107%; for organs at risk (OARs): bladder (mean < 45 Gy, D2% < 56 Gy, D30% < 35 Gy), femurs (D2% < 47 Gy), small bowel (mean < 30 Gy, D2% < 56 Gy). MU and delivery time scored treatment efficiency. Results: All techniques fulfilled objectives on maximum dose. Some deviations were observed on minimum dose for PTV. Uniformity (D5–D95) on PTVI resulted 6.6 ± 1.4% for IMRT and ranged from 5.7 ± 0.3% to 8.1 ± 0.8% for RA plans (±1 standard deviation). Conformity index (CI95%) was 1.3 ± 0.1 (IMRT) and 1.4 ± 0.1 (all RA techniques). Bladder: all techniques resulted equivalent above 40 Gy; V30Gy 57% for the double arcs, 61% for RA1 and 65% for IMRT. Femurs: maximum dose was of the order of 41–42 Gy for all RA plans and 45 Gy for IMRT. Small bowel: all techniques respected planning objectives. The number of computed MU/fraction was 1531 ± 206 (IMRT), 468 ± 95 (RA1), and 545 ± 80 (RA2) leading to differences in treatment time: 9.4 ± 1.7 min for IMRT vs. 1.1 ± 0.0 min for RA1 and 2.6 ± 0.0 min for double arcs. Conclusion: RapidArc showed improvements in organs at risk and healthy tissue sparing with uncompromised target coverage when double arcs are applied. Optimal results were also achieved anyway with IMRT plans. Ó 2009 Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology 92 (2009) 118–124
The aim of the present study was to investigate the potential clinical role for anal canal cancer patients of the RapidArc technique based on the volumetric intensity-modulated arc therapy when compared to conventional fixed beam IMRT. RapidArc (RA), based on an investigation from Otto [1], aims to improve conformal avoidance of treatments and to reduce the treatment time per fraction. The realisation of these objectives is achieved with a minimum number of modulated arcs and the usage of dose rate and gantry speed variations. RA is an evolution of the intensity-modulated arc therapy (IMAT) concept [2–5]. IMAT users
* Corresponding author. Address: Radiation Oncology Department, Oncology Institute of Southern Switzerland, 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.2008.12.020
clarified the benefit of modulated arcs for the most complex cases [6–9]. Volumetric IMAT, or VMAT, has already been investigated for other clinical cases [10–12] including prostate, small brain tumours and cervix uteri cancer. Since RapidArc should provide a natural evolution of IMRT and IMAT, it is of primary importance to assess performances on a theoretical basis before clinical application. Treatment of advanced anal canal carcinoma is a complex problem because of the shape of target volumes and the need of minimising the involvement of organs at risk such as bladder, small bowel and genitals. IMRT was investigated for anal canal and there is some evidence of a positive role of it in improving conformal avoidance [13,14]. Menkarios et al. [15] showed that pelvic IMRT could reduce the involvement of small bowel, bladder and genitals at medium doses as 30–40 Gy. The concept of simultaneous integrated boost (SIB) as described by Menkarios was adopted also
A. Clivio et al. / Radiotherapy and Oncology 92 (2009) 118–124
for the present publication with the same dose prescription of 59.4 and 49.5 Gy. IMAT [6,9] needs the application of multiple, coplanar or noncoplanar arcs to be beneficial against IMRT or conformal techniques. This has implications on treatment time as shown in [6,7] that reaches the order of 6–14 min depending on the tumour size and complexity (similarly for helical tomotherapy, average treatment time is 11 min [16], not stratified according to tumour location). Since RapidArc aims to apply a minimal number of arcs, the target of the present investigation was therefore to start an investigation of the trade-off between single and multiple arcs on realistic clinical cases.
Materials and methods Patient selection and planning objectives CT data for 10 patients, five males and five females, were selected for the comparison. All of them were affected by squamous cell carcinoma in the anal canal, T2–4, N0 or N+, M0 with the aim to be treated with curative intent. CT scans were acquired, supine position, with 3 mm slice thickness. The main organs at risk (OARs) were bladder and femurs. For the small bowel, the entire region encompassing it was outlined by the radiation oncologist. The genitals (both internal and external) were considered to be single organs (e.g. vulva, penis, and testis) as well as comprehensive global ‘organs’. The healthy tissue was defined by the patients volume covered by the CT scan minus the PTV. PTVs were defined from the clinical target volumes CTVs adding three-dimensional 8 mm margins. Two CTVs were defined: CTVI, including anal canal only, and CTVII including anal canal, internal iliac and inguinal lymph-nodes and pre-sacral region [17]. Dose prescription was assigned according to a SIB scheme with 59.4 Gy to the primary tumour (PTVI) in 1.8 Gy fractions and 49.5 Gy to the risk volume, including inguinal nodal stations (PTVII), in 33 fractions. All plans were normalized to the mean dose to the primary PTV (typical renormalization factors are inferior to 2–3%). For all PTVs, plans aimed to achieve minimum dose larger than 95% of the prescribed dose and a maximum lower than 107%. Planning objectives for OARs were defined as follows. Bladder: maximum dose (D2%) < 56 Gy, mean dose < 45 Gy, D30% < 35 Gy. Femurs: maximum dose (D2%) < 47 Gy. Small bowel: maximum dose (D2%) < 56 Gy, mean < 30 Gy. Planning techniques Three sets of plans were compared on the Varian Eclipse treatment planning system (TPS) in this study. 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 selection of a lower dose rate for IMRT (derived from clinical practice at our institute) does not affect optimisation, because the DR is not accounted for in that phase. On the contrary, for RA it is necessary to let the system free to use the entire range of achievable DRs to exploit the whole modulation phase space during optimisation. The Anisotropic Analytical Algorithm (AAA), which is the photon dose calculation algorithm, was used for all cases [18–21]. Plan optimisation was in both cases completely disentangled from the dose calculation and performed with the DVO (IMRT) and PRO (RapidArc) algorithms implemented in Eclipse, version 8.6.05. The dose calculation grid was set to 2.5 mm.
119
IMRT: Reference plans were computed as intensity modulation with fixed gantry and dynamic sliding window [22–24]. Plans were individually optimised using 7–9 coplanar fields (mean 8.4 ± 0.9 fields) selecting for each patient the best geometry. RapidArc (RA): To achieve the desired level of modulation required, the optimiser is enabled to continuously vary the instantaneous dose rate (DR), MLC leaf positions, as well as the gantry rotational speed. Details about the RapidArc process can be found in [10,11]. Four sets of plans were optimised and analysed. (i) RA1 consisting of a single 360° rotation with couch angle set to 0° and collimator set to either 30° or 36° (manually optimised to minimise maximum jaws openings). (ii) RA2 consisting of two coplanar arcs of 360° optimised simultaneously and to be delivered with opposite rotation (clockand counter-clock-wise). For RA2 plans the couch was set to 0° for both arcs, while the collimator rotation was set to the same angle X as in RA1 plans for the first arc and to 360° X for the second arc. The application of two arcs aims to increase the modulation factor during optimisation. In fact, because each individual arc is bound to 177 control points, the application of two independent arcs that are simultaneously optimised might allow the optimiser to achieve higher target homogeneity and lower OARs involvement than in other IMAT applications. Evaluation tools The 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. The homogeneity of the treatment was expressed in terms of 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%, 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 VxGy and DY% values. For healthy tissue, the integral dose, ‘‘DoseInt”, is defined as the integral of the absorbed dose extended 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. All data presented in the text, tables and figures refer to the mean and standard deviation averaged over the 10 cases. Average cumulative DVHs for PTV, OARs and healthy tissue were built from the individual DVHs averaging the corresponding volumes over the patient’s cohort each dose bin (0.02 Gy in this case). Paired, two-tails Student’s t-test was applied to the parameters of Table 1, and results were considered statistically significant for p < 0.05. To investigate the technical aspects of RA plans and appraise performances with respect to IMRT, MU/fraction has been scored. In addition, for each RA plan the dose rate, gantry speed and MU/° variations were derived per each arc and analysed in quantitative terms. Results Dose distributions for one example are shown in Fig. 1 for two axial views. Figs. 2 and 3 show the average DVHs for the PTVI, the
120
Volumetric modulater arc theraphy for anal canal cancer
Table 1 Summary of the dosimetric results for PTVI (59.4 Gy) and PTVII–PTVI (49.5 Gy), organs at risk and healthy tissue. Objective
IMRT
RA1
RA2
PTVI Dpres = 59.4 Gy, V = 347 ± 96 cm3 Mean D2% [%] D98% [%] D5–D95 [%]
100.0 <107 >95 Minimise
100.0 ± 0.0 103.8 ± 1.0 94.6 ± 1.3 6.6 ± 1.4
100.0 ± 0.0 105.0 ± 0.8* 94.6 ± 0.7 8.1 ± 0.8
100.0 ± 0.0 103.4 ± 0.5 95.9 ± 0.5* 5.7 ± 0.3
PTVII–PTVI Dpres = 49.5 Gy, V = 1307 ± 355 cm3 Mean D2% [%] D98% [%] D5–D95 [%] CI95
100.0 <107 >95 Minimise 1.0
102.7 ± 0.8 113.9 ± 3.6 95.1 ± 1.6 13.7 ± 3.8 1.03 ± 0.1
102.7 ± 0.9 114.9 ± 3.5 93.2 ± 1.6 16.4 ± 3.6 1.04 ± 0.1*
102.7 ± 0.7 114.0 ± 3.4 95.7 ± 1.0 13.4 ± 3.9 1.04 ± 0.1*
Bladder [120 ± 70 cm3] Mean [Gy] D2% [Gy] D67% [Gy] V30Gy [%]
<45 Gy <56 Gy Minimise Minimise
37.0 ± 7.3 54.7 ± 5.5 30.8 ± 10.2 64.9 ± 20.4
36.1 ± 8.2 55.4 ± 4.9 29.3 ± 11.6 60.7 ± 20.1
35.1 ± 7.9 55.2 ± 4.3 27.4 ± 11.9 57.4 ± 19.3
Femurs [153 ± 63 cm3] Mean [Gy] D2% [Gy] V45Gy [%]
Minimise <47 Gy Minimise
25.7 ± 2.8 44.9 ± 3.8 3.2 ± 2.9
21.9 ± 3.4* 42.4 ± 2.2* 0.8 ± 0.8*
21.0 ± 2.3* 41.4 ± 2.1* 0.5 ± 0.6*
Small bowel [2483 ± 774 cm3] Mean [Gy] D2% [Gy] V30Gy [%]
<30 Gy <56 Gy Minimise
20.5 ± 6.4 51.6 ± 0.6 33.1 ± 12.7
21.3 ± 5.9 52.4 ± 0.8 34.8 ± 11.4
21.2 ± 5.6 51.5 ± 0.8 34.0 ± 10.2
Healthy tissue Mean [Gy] V10Gy [%] DoseInt [Gy * cm3 * 10
Minimise Minimise –
15.7 ± 1.7 52.0 ± 6.5 2.89 ± 0.78
15.9 ± 2.4 53.5 ± 8.3 2.87 ± 0.72
15.9 ± 2.2 54.7 ± 8.4 2.88 ± 0.74
5
]
Statistical significance (p < 0.05) is reported (*) against IMRT benchmark from paired t-test analysis.
PTVII–PTVI, organs at risk and healthy tissue. In Table 1, the numerical findings from DVH analysis on PTV and main OARs are reported. In Table 2, the findings on monitor units and delivery parameters are summarised. Data are presented as averages over the investigated patients and errors indicated inter-patient variability at one standard deviation level. Target coverage and dose homogeneity Data in the tables are reported in the percentage values of the dose prescribed to each PTV (e.g. for PTVII–PTVI 100% corresponds to 49.5 Gy) while in the figures graphs are expressed in the percentage of the dose prescribed to PTVI, i.e. 59.4 Gy. All techniques result in similar target coverage. D98% has reached values higher than 95% for double arc RA plans. Organs at risk Bladder: All techniques respected planning objectives on mean and maximum significant dose, while the objective on D30% was respected only for the portion of bladder not included inside PTVII. The volume of bladder irradiated at medium–low dose levels (between 10 and 40 Gy) was reduced with double RA arcs as it is shown by the V30Gy parameter. Femurs: D2% resulted inferior to the objective of 47 Gy for all techniques; RA plans allow a statistically significant additional reduction compared to IMRT, ranging from 2.5 to 3.6 Gy. Small bowel: Planning objectives were met by all techniques and no relevant difference was observed between RA and IMRT.
Genitals (male): In the male case, prostate and seminal vesicles were either entirely or largely included in PTVII and no sparing effort was devoted to these organs. Attention was therefore devoted to external genitals (penis including bulb and testis). Testes were almost entirely spared and involved only by scattering with a mean dose of: 4.6 ± 4.4, 7.0 ± 4.9, 5.7 ± 2.9 Gy for IMRT, RA1 and RA2, respectively. D2% was 9.6 ± 10.3, 11.4 ± 8.9, and 9.4 ± 6.0, respectively. Concerning penis, mean dose was 12.8 ± 2.0, 14.0 ± 4.5, and 10.4 ± 1.8. From mean DVH graphs, it can be noticed that double arcs are superior to single arc and IMRT. Genitals (female): Uterus, ovaries and vagina are either fully or largely included inside PTVII, while the vulva is normally entirely outside the irradiated volume. Mean dose to vulva was 18.2 ± 1.8, 21.8 ± 4.5, and 18.1 ± 2.0 Gy, respectively, for IMRT, RA1, and RA2, while the mean dose for the uninvolved genitals was 21.1 ± 1.2, 25.1 ± 5.7, and 21.8 ± 2.0 Gy, respectively. In this case, IMRT and double arc RA are equivalent (from average DVH graphs), while RA1 is inferior in the male case with maximum differences as large as 10 Gy in D30% for the vulva. Healthy tissue: The planning objectives for healthy tissue were not formalised in numerical terms, but the strategy was to minimise its involvement. RA1, RA2 and IMRT presented similar shapes in the DVH of the healthy tissue reflecting equivalent numerical findings. None of the parameters investigated here resulted to be statistically different between techniques. To complement investigation on the healthy tissue, mean dose on peripheral shells at 5, 10 cm from the PTVII surface was scored. These shells were defined as 2 mm thick walls expanding in 3D the PTVII as described in [10]. Mean dose at 5 cm from PTVII resulted 18.0 ± 2.7%, 17.8 ± 2.5%, and 17.9 ± 2.4% of the prescribed dose for IMRT, RA1, and RA2. At 10 cm mean dose was 8.5 ± 5.4%, 6.5 ± 4.3%, and 6.6 ± 4.5%, respectively.
A. Clivio et al. / Radiotherapy and Oncology 92 (2009) 118–124
121
Fig. 1. Dose distributions on two axial views for one representative case.
Monitor units, delivery time and technical features The mean number of MU per fraction, the mean dose rate (DR), gantry speed (GS), MU/° and treatment time are reported in Table 2, averaged over the two arcs in the case of RA2 plans. IMRT presented values of MU/fraction roughly three times higher with respect to RapidArc. To note that for RA, all cases except one single patient in RA1, resulted in constant gantry speed rotating at the maximum speed allowed (5.5°/s). Treatment time is defined as the beam on time plus the time needed to re-configure the linac between one field (or arc) and the other. All treatment times were measured at linac during simulated delivery. It has to be considered that all IMRT fields in this study are wider than 15 cm in the direction of MLC motion and this implied that all were split into two carriage sequences due to the current HW limitations of the system, meaning that for each gantry position, two IM fields shall be delivered. On the contrary, for RA with double arcs, the re-configuration of the linac is relatively fast because the only ‘‘slow” element is given by the collimator rotation taking about 25 s and all motions can be managed remotely. Gantry rotation in clock and counter-clock directions for the two arcs contributes to minimise the reset time. Discussion This study reports a comparison of the volumetric-modulated arc therapy, RapidArc technique, with single or double coplanar arcs against fixed beam IMRT for anal canal cancer patients. Given the relatively limited literature on the application of IMRT, there is no wide experience for achievable results in terms of conformal avoidance. A qualitative comparison of the IMRT findings from the present investigation to the data published by Menkarios
et al. [15] shows that the benchmark adopted to appraise RapidArc performances is quite extreme. In fact, considering the most favourable scenario of [15], V30Gy for the bladder ranged from 77% to 89%, while our IMRT achieved a value of 65%, from 12% to 24% lower; similarly, V45Gy for the femoral heads ranged from 9% to 16% in [15] and was 3% in this analysis; V30Gy for small bowel was reduced from 43–45% to 33%; considering genitals, V30Gy data reported in [15] ranged from 33% to 64%, while in this analysis the same quantity for female patients was 18% and only 1% for male patients. Mean dose to genitals reported by Milano et al. [14] is of 30 Gy, while in our study it was 10 Gy for males and 18 Gy for females. To conclude the comparison, it is obvious that IMRT is an excellent method for treating anal canal and that the data of this study largely improve previously published findings. Compared to this high quality IMRT, RapidArc allows to improve organs at risk sparing at various levels, whereas with the double arc method, it also allows to moderately improve target coverage and homogeneity. Double arc plans are manifested to be more robust in all respects compared to single arc. IMRT achieved better conformality than RA and this is also reflected in the healthy tissue (e.g. V10Gy, although without statistical significance). Concerning normalization, the strategy followed in this study induces some cold spots in the target. A different strategy, e.g. normalization aiming to cover a desired proportion of PTV volume with a desired minimum dose, although not fully compatible with current trends and ICRU recommendations could be clinically preferable. Nevertheless this latter approach might generate a general overdose simply because fixing the dose normalization on (low) dose levels represents few pixels. In summary, from the dosimetrical point of view (i) all techniques, including IMRT, met in general or largely improved, planning objectives; (ii) differences between single and double arc
122
Volumetric modulater arc theraphy for anal canal cancer
Fig. 2. Mean DVHs for PTVI, PTVII–PTVI, bladder, femurs, small bowel and healthy tissue.
were not statistically significant; (iii) the clinical relevance of the observed differences between techniques is probably not of concern. It will be fundamental to establish proper clinical protocols to ascertain the measured clinical toxicity against expectations based on the dosimetric investigations. No differences were observed between RapidArc and IMRT in terms of healthy tissue, in contrast with what is reported in [10,11]. This is explained by the fact that, in the anal canal, the target is quite superficial and it spread over a wide body volume which is composed by the tumour bed and the two well-separated nodal stations with organs at risk outside (e.g. femurs) and inside (e.g. bladder and genitals) the envelope.
One of the major issues that should be clarified for volumetricmodulated arc therapy is whether single or multiple arcs shall be applied to realise proper treatments. It was shown [6–9] that several arcs were necessary to generate modulation patterns with multiple intensity levels. In this study, we appraised the problem of arc multiplicity investigating single and double arcs. Results showed that double arcs can slightly improve the sparing of organs at risk. The usage of two arcs with RA, intuitively, shall guarantee an higher freedom in dose modulation compared to single arc treatments because of two factors: (i) RA is bound to describe one single arc with a sequence of 177 control points; (ii) the optimisation of two arcs, although simultaneous, is independent, i.e.
A. Clivio et al. / Radiotherapy and Oncology 92 (2009) 118–124
123
Fig. 3. Mean dose volume histograms for genitals (male and female), penis and vulva.
each arc can create a completely unrelated sequence of MLC shapes, dose rates and gantry speed combinations. As a consequence, when necessary for dosimetric reasons (as in the present due to target-OAR geometrical complexity), two arcs might lead to better dose distributions, as it is shown here, also against an hypothetic single arc with 360 control points. The trade-off between the number of arcs and the quality of planing is a prolonged treatment time. Literature results from other IMAT approaches [6,7,16] reports mean delivery time ranging from 6.3 to 14 min for C-arm linac-based techniques to 11 min for helical tomotherapy depending on the clinical indication. These values are consistent with the 9–10 min of pure treat-
Table 2 Summary of the MU/fraction and of delivery parameters.
ment time (without any verification imaging process) measured for the IMRT plans of this study and shall be compared to the 1.1 min that are sufficient for single arc and only about 2.6 min necessary for double arcs with RA. In terms of MU/fraction, it is noticeable that RA2 showed some 12% lower output compared to RA1. In addition, at the given dose prescription the mean dose rate and the number of MU/° are obviously roughly inversely proportional to the number of arcs applied. This means that beyond two arcs the linac should be operated at very low efficiency levels (the minimum DR of double arcs is as low as 70 MU/min) and eventually could reach the limit of too low MU/° that cannot be compensated by faster rotation of the gantry. To conclude, the trade-off between single and double arc from the delivery point of view is expressed by a longer treatment time and a less efficient usage of the accelerated beam with very low dose rates and low angular delivery. Concerning MU and delivery times with IMRT, the value we report is relative to the sliding window technique. It is possible that different approaches might lead to lower values. Similarly, higher dose rate for IMRT would linearly reduce the beam on time but not the effective treatment time, which is dominated by the beam multiplicity and the dead times necessary to move from one field to the next.
IMRT
RA1
RA2
MU/fraction Ratioa
1531 ± 206 1
468 ± 95 0.31 ± 0.07
545 ± 80 0.36 ± 0.07
DR mean [MU/min]
300
GS mean [°/s]
N/A
428.1 ± 90.4 [231.6, 600.0] 5.5 ± 0.2 [4.4, 5.5]
252.9 ± 72.9 [89.7, 538.1] 5.5 ± 0.0 [5.5, 5.5]
MU/° mean
N/A
1.3 ± 0.3 [0.7, 2.3]
0.8 ± 0.2 [0.3, 1.6]
Conclusions
Delivery time [min]
9.4 ± 1.7
1.1 ± 0.0
2.6 ± 0.0b
RapidArc was investigated for anal canal cancer in comparison to IMRT. Although excellent dosimetric results were achieved with IMRT, RapidArc with double arc allowed to add improvement in sparing some organs at risk and uncompromised target coverage.
a
With respect to MU/fraction of IMRT. b Incorporates 25 s fixed time to re-configure linac for second arc, including couch and collimator rotation to start position.
124
Volumetric modulater arc theraphy for anal canal cancer
Single arcs resulted to be slightly inferior, on target coverage and on sparing some OARs. Double arcs are acceptable also in terms of treatment time since, although longer with respect to single arc, it results 3.5 times shorter than average treatment time for IMRT. Clinical investigations should be performed to assess the actual benefit in terms of reduced toxicity or improved control shown in the physical planning studies. Disclosure Dr. L. Cozzi acts as Scientific Advisor to Varian Medical Systems and as Head of Research and Technological Development to Oncology Institute of Southern Switzerland, IOSI, Bellinzona. Acknowledgement The current investigation was partially covered by a Varian research grant to IOSI. References [1] Otto K. Volumetric modulated arc therapy: IMRT in a single arc. Med Phys 2008;35:310–7. [2] Cameron C. Sweeping-window arc therapy: an implementation of rotational IMRT with automatic beam-weight calculation. Phys Med Biol 2005;50:4317–36. [3] Cotrutz C, Kappas C, Webb S. Intensity modulated arc therapy (IMAT) with centrally blocked rotational fields. Phys Med Biol 2000;45:2185–206. [4] 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. [5] 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. [6] 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. [7] Duthoy W, De Gersem W, Vergote K, Coghe M, Botenberf T, De Deene Y, et al. Whole abdominal radiotherapy (WAPRT) using intensity modulated arc therapy (IMAT): first clinical experience. Int J Radiat Oncol Biol Phys 2003;57:1019–32. [8] Yu CX. Intensity-modulated arc therapy with dynamic multileaf collimation: an alternative to tomotherapy. Phys Med Biol 1995;40:1435–49. [9] 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.
[10] 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. [11] 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;89:254–62. [12] 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 three-dimensional conformal radiotherapy. Int J Radiat Oncol Biol Phys 2008;72:996–1001. [13] Chen YJ, Liu A, Tsai P, Vora N, Pezner R, Shultheiss T, et al. Organ sparing by conformal avoidance intensity modulated radiation therapy for anal cancer: dosimetric evaluation of coverage of pelvis and inguinal/femoral nodes. Int J Radiat Oncol Biol Phys 2005;63:274–81. [14] Milano M, Jani A, Farrey K, Rash C, Heimann R, Chmura S. Intensity modulated radiation therapy in the treatment of anal cancer: toxicity and clinical outcome. Int J Radiat Oncol Biol Phys 2005;63:354–61. [15] Menkarios C, Azria D, Laliberte B, Llacer Moscardo C, Gourgou S, Lemanski C, et al. Optimal organ sparing intensity modulated radiation therapy regimen for the treatment of locally advanced anal canal carcinoma: a comparison of conventional and IMRT plans. Radiat Oncol 2007;2:41. [16] 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. [17] Bosset J, Roelofsen F, Morgan D, Budach V, Coucke P, Jager J, et al. Shortened irradiation scheme, continuous infusion of 5-fluorouracil and fractionation of mitomycin C in locally advanced anal carcinomas. Results of a phase II study of the European organization for Research and Treatment of Cancer. Radiotherapy and Gastrointestinal Cooperative Groups. Eur J Cancer 2003;39:45–51. [18] 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. [19] 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. [20] 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. [21] 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. [22] 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. [23] Convery D, Rosenbloom M. The generation of intensity-modulated fields for conformal radiotherapy by dynamic collimation. Phys Med Biol 1992;37:1359–74. [24] Spirou S, Chui C. A gradient inverse planning algorithm with dose-volume constrains. Med Phys 1998;25:321–33.
Contents lists available at ScienceDirect
Radiotherapy and Oncology journal homepage: www.thegreenjournal.com
Anal cancer IMRT
Volumetric-modulated arc radiotherapy for carcinomas of the anal canal: A treatment planning comparison with fixed field IMRT Alessandro Clivio a, Antonella Fogliata a, Alessandra Franzetti-Pellanda b, Giorgia Nicolini a, Eugenio Vanetti a, Rolf Wyttenbach c,d, Luca Cozzi a,* a
Medical Physics, Oncology Institute of Southern Switzerland, Bellinzona, Switzerland Radiation Oncology Department, Oncology Institute of Southern Switzerland, Bellinzona, Switzerland c Ospedale San Giovanni, Radiology, Bellinzona, Switzerland d University of Bern, Bern, Switzerland b
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 30 January 2009 Keywords: RapidArc IMRT Anal canal radiation therapy Volumetric-modulated arc therapy
a b s t r a c t Purpose: A treatment planning study was performed to compare volumetric-modulated arc radiotherapy against conventional fixed field IMRT. Materials and methods: CT datasets of 10 patients affected by carcinoma of the anal canal were included and five plans were generated for each case: fixed beam IMRT, single (RA1)- and double (RA2)-modulated arcs with the RapidArc technique. Dose prescription was set according to a simultaneous integrated boost strategy to 59.4 Gy to the primary tumour PTVI (at 1.8 Gy/fraction) and to 49.5 Gy to risk area including inguinal nodes, PTVII. Planning objectives for PTV were minimum dose >95%, maximum dose < 107%; for organs at risk (OARs): bladder (mean < 45 Gy, D2% < 56 Gy, D30% < 35 Gy), femurs (D2% < 47 Gy), small bowel (mean < 30 Gy, D2% < 56 Gy). MU and delivery time scored treatment efficiency. Results: All techniques fulfilled objectives on maximum dose. Some deviations were observed on minimum dose for PTV. Uniformity (D5–D95) on PTVI resulted 6.6 ± 1.4% for IMRT and ranged from 5.7 ± 0.3% to 8.1 ± 0.8% for RA plans (±1 standard deviation). Conformity index (CI95%) was 1.3 ± 0.1 (IMRT) and 1.4 ± 0.1 (all RA techniques). Bladder: all techniques resulted equivalent above 40 Gy; V30Gy 57% for the double arcs, 61% for RA1 and 65% for IMRT. Femurs: maximum dose was of the order of 41–42 Gy for all RA plans and 45 Gy for IMRT. Small bowel: all techniques respected planning objectives. The number of computed MU/fraction was 1531 ± 206 (IMRT), 468 ± 95 (RA1), and 545 ± 80 (RA2) leading to differences in treatment time: 9.4 ± 1.7 min for IMRT vs. 1.1 ± 0.0 min for RA1 and 2.6 ± 0.0 min for double arcs. Conclusion: RapidArc showed improvements in organs at risk and healthy tissue sparing with uncompromised target coverage when double arcs are applied. Optimal results were also achieved anyway with IMRT plans. Ó 2009 Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology 92 (2009) 118–124
The aim of the present study was to investigate the potential clinical role for anal canal cancer patients of the RapidArc technique based on the volumetric intensity-modulated arc therapy when compared to conventional fixed beam IMRT. RapidArc (RA), based on an investigation from Otto [1], aims to improve conformal avoidance of treatments and to reduce the treatment time per fraction. The realisation of these objectives is achieved with a minimum number of modulated arcs and the usage of dose rate and gantry speed variations. RA is an evolution of the intensity-modulated arc therapy (IMAT) concept [2–5]. IMAT users
* Corresponding author. Address: Radiation Oncology Department, Oncology Institute of Southern Switzerland, 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.2008.12.020
clarified the benefit of modulated arcs for the most complex cases [6–9]. Volumetric IMAT, or VMAT, has already been investigated for other clinical cases [10–12] including prostate, small brain tumours and cervix uteri cancer. Since RapidArc should provide a natural evolution of IMRT and IMAT, it is of primary importance to assess performances on a theoretical basis before clinical application. Treatment of advanced anal canal carcinoma is a complex problem because of the shape of target volumes and the need of minimising the involvement of organs at risk such as bladder, small bowel and genitals. IMRT was investigated for anal canal and there is some evidence of a positive role of it in improving conformal avoidance [13,14]. Menkarios et al. [15] showed that pelvic IMRT could reduce the involvement of small bowel, bladder and genitals at medium doses as 30–40 Gy. The concept of simultaneous integrated boost (SIB) as described by Menkarios was adopted also
A. Clivio et al. / Radiotherapy and Oncology 92 (2009) 118–124
for the present publication with the same dose prescription of 59.4 and 49.5 Gy. IMAT [6,9] needs the application of multiple, coplanar or noncoplanar arcs to be beneficial against IMRT or conformal techniques. This has implications on treatment time as shown in [6,7] that reaches the order of 6–14 min depending on the tumour size and complexity (similarly for helical tomotherapy, average treatment time is 11 min [16], not stratified according to tumour location). Since RapidArc aims to apply a minimal number of arcs, the target of the present investigation was therefore to start an investigation of the trade-off between single and multiple arcs on realistic clinical cases.
Materials and methods Patient selection and planning objectives CT data for 10 patients, five males and five females, were selected for the comparison. All of them were affected by squamous cell carcinoma in the anal canal, T2–4, N0 or N+, M0 with the aim to be treated with curative intent. CT scans were acquired, supine position, with 3 mm slice thickness. The main organs at risk (OARs) were bladder and femurs. For the small bowel, the entire region encompassing it was outlined by the radiation oncologist. The genitals (both internal and external) were considered to be single organs (e.g. vulva, penis, and testis) as well as comprehensive global ‘organs’. The healthy tissue was defined by the patients volume covered by the CT scan minus the PTV. PTVs were defined from the clinical target volumes CTVs adding three-dimensional 8 mm margins. Two CTVs were defined: CTVI, including anal canal only, and CTVII including anal canal, internal iliac and inguinal lymph-nodes and pre-sacral region [17]. Dose prescription was assigned according to a SIB scheme with 59.4 Gy to the primary tumour (PTVI) in 1.8 Gy fractions and 49.5 Gy to the risk volume, including inguinal nodal stations (PTVII), in 33 fractions. All plans were normalized to the mean dose to the primary PTV (typical renormalization factors are inferior to 2–3%). For all PTVs, plans aimed to achieve minimum dose larger than 95% of the prescribed dose and a maximum lower than 107%. Planning objectives for OARs were defined as follows. Bladder: maximum dose (D2%) < 56 Gy, mean dose < 45 Gy, D30% < 35 Gy. Femurs: maximum dose (D2%) < 47 Gy. Small bowel: maximum dose (D2%) < 56 Gy, mean < 30 Gy. Planning techniques Three sets of plans were compared on the Varian Eclipse treatment planning system (TPS) in this study. 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 selection of a lower dose rate for IMRT (derived from clinical practice at our institute) does not affect optimisation, because the DR is not accounted for in that phase. On the contrary, for RA it is necessary to let the system free to use the entire range of achievable DRs to exploit the whole modulation phase space during optimisation. The Anisotropic Analytical Algorithm (AAA), which is the photon dose calculation algorithm, was used for all cases [18–21]. Plan optimisation was in both cases completely disentangled from the dose calculation and performed with the DVO (IMRT) and PRO (RapidArc) algorithms implemented in Eclipse, version 8.6.05. The dose calculation grid was set to 2.5 mm.
119
IMRT: Reference plans were computed as intensity modulation with fixed gantry and dynamic sliding window [22–24]. Plans were individually optimised using 7–9 coplanar fields (mean 8.4 ± 0.9 fields) selecting for each patient the best geometry. RapidArc (RA): To achieve the desired level of modulation required, the optimiser is enabled to continuously vary the instantaneous dose rate (DR), MLC leaf positions, as well as the gantry rotational speed. Details about the RapidArc process can be found in [10,11]. Four sets of plans were optimised and analysed. (i) RA1 consisting of a single 360° rotation with couch angle set to 0° and collimator set to either 30° or 36° (manually optimised to minimise maximum jaws openings). (ii) RA2 consisting of two coplanar arcs of 360° optimised simultaneously and to be delivered with opposite rotation (clockand counter-clock-wise). For RA2 plans the couch was set to 0° for both arcs, while the collimator rotation was set to the same angle X as in RA1 plans for the first arc and to 360° X for the second arc. The application of two arcs aims to increase the modulation factor during optimisation. In fact, because each individual arc is bound to 177 control points, the application of two independent arcs that are simultaneously optimised might allow the optimiser to achieve higher target homogeneity and lower OARs involvement than in other IMAT applications. Evaluation tools The 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. The homogeneity of the treatment was expressed in terms of 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%, 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 VxGy and DY% values. For healthy tissue, the integral dose, ‘‘DoseInt”, is defined as the integral of the absorbed dose extended 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. All data presented in the text, tables and figures refer to the mean and standard deviation averaged over the 10 cases. Average cumulative DVHs for PTV, OARs and healthy tissue were built from the individual DVHs averaging the corresponding volumes over the patient’s cohort each dose bin (0.02 Gy in this case). Paired, two-tails Student’s t-test was applied to the parameters of Table 1, and results were considered statistically significant for p < 0.05. To investigate the technical aspects of RA plans and appraise performances with respect to IMRT, MU/fraction has been scored. In addition, for each RA plan the dose rate, gantry speed and MU/° variations were derived per each arc and analysed in quantitative terms. Results Dose distributions for one example are shown in Fig. 1 for two axial views. Figs. 2 and 3 show the average DVHs for the PTVI, the
120
Volumetric modulater arc theraphy for anal canal cancer
Table 1 Summary of the dosimetric results for PTVI (59.4 Gy) and PTVII–PTVI (49.5 Gy), organs at risk and healthy tissue. Objective
IMRT
RA1
RA2
PTVI Dpres = 59.4 Gy, V = 347 ± 96 cm3 Mean D2% [%] D98% [%] D5–D95 [%]
100.0 <107 >95 Minimise
100.0 ± 0.0 103.8 ± 1.0 94.6 ± 1.3 6.6 ± 1.4
100.0 ± 0.0 105.0 ± 0.8* 94.6 ± 0.7 8.1 ± 0.8
100.0 ± 0.0 103.4 ± 0.5 95.9 ± 0.5* 5.7 ± 0.3
PTVII–PTVI Dpres = 49.5 Gy, V = 1307 ± 355 cm3 Mean D2% [%] D98% [%] D5–D95 [%] CI95
100.0 <107 >95 Minimise 1.0
102.7 ± 0.8 113.9 ± 3.6 95.1 ± 1.6 13.7 ± 3.8 1.03 ± 0.1
102.7 ± 0.9 114.9 ± 3.5 93.2 ± 1.6 16.4 ± 3.6 1.04 ± 0.1*
102.7 ± 0.7 114.0 ± 3.4 95.7 ± 1.0 13.4 ± 3.9 1.04 ± 0.1*
Bladder [120 ± 70 cm3] Mean [Gy] D2% [Gy] D67% [Gy] V30Gy [%]
<45 Gy <56 Gy Minimise Minimise
37.0 ± 7.3 54.7 ± 5.5 30.8 ± 10.2 64.9 ± 20.4
36.1 ± 8.2 55.4 ± 4.9 29.3 ± 11.6 60.7 ± 20.1
35.1 ± 7.9 55.2 ± 4.3 27.4 ± 11.9 57.4 ± 19.3
Femurs [153 ± 63 cm3] Mean [Gy] D2% [Gy] V45Gy [%]
Minimise <47 Gy Minimise
25.7 ± 2.8 44.9 ± 3.8 3.2 ± 2.9
21.9 ± 3.4* 42.4 ± 2.2* 0.8 ± 0.8*
21.0 ± 2.3* 41.4 ± 2.1* 0.5 ± 0.6*
Small bowel [2483 ± 774 cm3] Mean [Gy] D2% [Gy] V30Gy [%]
<30 Gy <56 Gy Minimise
20.5 ± 6.4 51.6 ± 0.6 33.1 ± 12.7
21.3 ± 5.9 52.4 ± 0.8 34.8 ± 11.4
21.2 ± 5.6 51.5 ± 0.8 34.0 ± 10.2
Healthy tissue Mean [Gy] V10Gy [%] DoseInt [Gy * cm3 * 10
Minimise Minimise –
15.7 ± 1.7 52.0 ± 6.5 2.89 ± 0.78
15.9 ± 2.4 53.5 ± 8.3 2.87 ± 0.72
15.9 ± 2.2 54.7 ± 8.4 2.88 ± 0.74
5
]
Statistical significance (p < 0.05) is reported (*) against IMRT benchmark from paired t-test analysis.
PTVII–PTVI, organs at risk and healthy tissue. In Table 1, the numerical findings from DVH analysis on PTV and main OARs are reported. In Table 2, the findings on monitor units and delivery parameters are summarised. Data are presented as averages over the investigated patients and errors indicated inter-patient variability at one standard deviation level. Target coverage and dose homogeneity Data in the tables are reported in the percentage values of the dose prescribed to each PTV (e.g. for PTVII–PTVI 100% corresponds to 49.5 Gy) while in the figures graphs are expressed in the percentage of the dose prescribed to PTVI, i.e. 59.4 Gy. All techniques result in similar target coverage. D98% has reached values higher than 95% for double arc RA plans. Organs at risk Bladder: All techniques respected planning objectives on mean and maximum significant dose, while the objective on D30% was respected only for the portion of bladder not included inside PTVII. The volume of bladder irradiated at medium–low dose levels (between 10 and 40 Gy) was reduced with double RA arcs as it is shown by the V30Gy parameter. Femurs: D2% resulted inferior to the objective of 47 Gy for all techniques; RA plans allow a statistically significant additional reduction compared to IMRT, ranging from 2.5 to 3.6 Gy. Small bowel: Planning objectives were met by all techniques and no relevant difference was observed between RA and IMRT.
Genitals (male): In the male case, prostate and seminal vesicles were either entirely or largely included in PTVII and no sparing effort was devoted to these organs. Attention was therefore devoted to external genitals (penis including bulb and testis). Testes were almost entirely spared and involved only by scattering with a mean dose of: 4.6 ± 4.4, 7.0 ± 4.9, 5.7 ± 2.9 Gy for IMRT, RA1 and RA2, respectively. D2% was 9.6 ± 10.3, 11.4 ± 8.9, and 9.4 ± 6.0, respectively. Concerning penis, mean dose was 12.8 ± 2.0, 14.0 ± 4.5, and 10.4 ± 1.8. From mean DVH graphs, it can be noticed that double arcs are superior to single arc and IMRT. Genitals (female): Uterus, ovaries and vagina are either fully or largely included inside PTVII, while the vulva is normally entirely outside the irradiated volume. Mean dose to vulva was 18.2 ± 1.8, 21.8 ± 4.5, and 18.1 ± 2.0 Gy, respectively, for IMRT, RA1, and RA2, while the mean dose for the uninvolved genitals was 21.1 ± 1.2, 25.1 ± 5.7, and 21.8 ± 2.0 Gy, respectively. In this case, IMRT and double arc RA are equivalent (from average DVH graphs), while RA1 is inferior in the male case with maximum differences as large as 10 Gy in D30% for the vulva. Healthy tissue: The planning objectives for healthy tissue were not formalised in numerical terms, but the strategy was to minimise its involvement. RA1, RA2 and IMRT presented similar shapes in the DVH of the healthy tissue reflecting equivalent numerical findings. None of the parameters investigated here resulted to be statistically different between techniques. To complement investigation on the healthy tissue, mean dose on peripheral shells at 5, 10 cm from the PTVII surface was scored. These shells were defined as 2 mm thick walls expanding in 3D the PTVII as described in [10]. Mean dose at 5 cm from PTVII resulted 18.0 ± 2.7%, 17.8 ± 2.5%, and 17.9 ± 2.4% of the prescribed dose for IMRT, RA1, and RA2. At 10 cm mean dose was 8.5 ± 5.4%, 6.5 ± 4.3%, and 6.6 ± 4.5%, respectively.
A. Clivio et al. / Radiotherapy and Oncology 92 (2009) 118–124
121
Fig. 1. Dose distributions on two axial views for one representative case.
Monitor units, delivery time and technical features The mean number of MU per fraction, the mean dose rate (DR), gantry speed (GS), MU/° and treatment time are reported in Table 2, averaged over the two arcs in the case of RA2 plans. IMRT presented values of MU/fraction roughly three times higher with respect to RapidArc. To note that for RA, all cases except one single patient in RA1, resulted in constant gantry speed rotating at the maximum speed allowed (5.5°/s). Treatment time is defined as the beam on time plus the time needed to re-configure the linac between one field (or arc) and the other. All treatment times were measured at linac during simulated delivery. It has to be considered that all IMRT fields in this study are wider than 15 cm in the direction of MLC motion and this implied that all were split into two carriage sequences due to the current HW limitations of the system, meaning that for each gantry position, two IM fields shall be delivered. On the contrary, for RA with double arcs, the re-configuration of the linac is relatively fast because the only ‘‘slow” element is given by the collimator rotation taking about 25 s and all motions can be managed remotely. Gantry rotation in clock and counter-clock directions for the two arcs contributes to minimise the reset time. Discussion This study reports a comparison of the volumetric-modulated arc therapy, RapidArc technique, with single or double coplanar arcs against fixed beam IMRT for anal canal cancer patients. Given the relatively limited literature on the application of IMRT, there is no wide experience for achievable results in terms of conformal avoidance. A qualitative comparison of the IMRT findings from the present investigation to the data published by Menkarios
et al. [15] shows that the benchmark adopted to appraise RapidArc performances is quite extreme. In fact, considering the most favourable scenario of [15], V30Gy for the bladder ranged from 77% to 89%, while our IMRT achieved a value of 65%, from 12% to 24% lower; similarly, V45Gy for the femoral heads ranged from 9% to 16% in [15] and was 3% in this analysis; V30Gy for small bowel was reduced from 43–45% to 33%; considering genitals, V30Gy data reported in [15] ranged from 33% to 64%, while in this analysis the same quantity for female patients was 18% and only 1% for male patients. Mean dose to genitals reported by Milano et al. [14] is of 30 Gy, while in our study it was 10 Gy for males and 18 Gy for females. To conclude the comparison, it is obvious that IMRT is an excellent method for treating anal canal and that the data of this study largely improve previously published findings. Compared to this high quality IMRT, RapidArc allows to improve organs at risk sparing at various levels, whereas with the double arc method, it also allows to moderately improve target coverage and homogeneity. Double arc plans are manifested to be more robust in all respects compared to single arc. IMRT achieved better conformality than RA and this is also reflected in the healthy tissue (e.g. V10Gy, although without statistical significance). Concerning normalization, the strategy followed in this study induces some cold spots in the target. A different strategy, e.g. normalization aiming to cover a desired proportion of PTV volume with a desired minimum dose, although not fully compatible with current trends and ICRU recommendations could be clinically preferable. Nevertheless this latter approach might generate a general overdose simply because fixing the dose normalization on (low) dose levels represents few pixels. In summary, from the dosimetrical point of view (i) all techniques, including IMRT, met in general or largely improved, planning objectives; (ii) differences between single and double arc
122
Volumetric modulater arc theraphy for anal canal cancer
Fig. 2. Mean DVHs for PTVI, PTVII–PTVI, bladder, femurs, small bowel and healthy tissue.
were not statistically significant; (iii) the clinical relevance of the observed differences between techniques is probably not of concern. It will be fundamental to establish proper clinical protocols to ascertain the measured clinical toxicity against expectations based on the dosimetric investigations. No differences were observed between RapidArc and IMRT in terms of healthy tissue, in contrast with what is reported in [10,11]. This is explained by the fact that, in the anal canal, the target is quite superficial and it spread over a wide body volume which is composed by the tumour bed and the two well-separated nodal stations with organs at risk outside (e.g. femurs) and inside (e.g. bladder and genitals) the envelope.
One of the major issues that should be clarified for volumetricmodulated arc therapy is whether single or multiple arcs shall be applied to realise proper treatments. It was shown [6–9] that several arcs were necessary to generate modulation patterns with multiple intensity levels. In this study, we appraised the problem of arc multiplicity investigating single and double arcs. Results showed that double arcs can slightly improve the sparing of organs at risk. The usage of two arcs with RA, intuitively, shall guarantee an higher freedom in dose modulation compared to single arc treatments because of two factors: (i) RA is bound to describe one single arc with a sequence of 177 control points; (ii) the optimisation of two arcs, although simultaneous, is independent, i.e.
A. Clivio et al. / Radiotherapy and Oncology 92 (2009) 118–124
123
Fig. 3. Mean dose volume histograms for genitals (male and female), penis and vulva.
each arc can create a completely unrelated sequence of MLC shapes, dose rates and gantry speed combinations. As a consequence, when necessary for dosimetric reasons (as in the present due to target-OAR geometrical complexity), two arcs might lead to better dose distributions, as it is shown here, also against an hypothetic single arc with 360 control points. The trade-off between the number of arcs and the quality of planing is a prolonged treatment time. Literature results from other IMAT approaches [6,7,16] reports mean delivery time ranging from 6.3 to 14 min for C-arm linac-based techniques to 11 min for helical tomotherapy depending on the clinical indication. These values are consistent with the 9–10 min of pure treat-
Table 2 Summary of the MU/fraction and of delivery parameters.
ment time (without any verification imaging process) measured for the IMRT plans of this study and shall be compared to the 1.1 min that are sufficient for single arc and only about 2.6 min necessary for double arcs with RA. In terms of MU/fraction, it is noticeable that RA2 showed some 12% lower output compared to RA1. In addition, at the given dose prescription the mean dose rate and the number of MU/° are obviously roughly inversely proportional to the number of arcs applied. This means that beyond two arcs the linac should be operated at very low efficiency levels (the minimum DR of double arcs is as low as 70 MU/min) and eventually could reach the limit of too low MU/° that cannot be compensated by faster rotation of the gantry. To conclude, the trade-off between single and double arc from the delivery point of view is expressed by a longer treatment time and a less efficient usage of the accelerated beam with very low dose rates and low angular delivery. Concerning MU and delivery times with IMRT, the value we report is relative to the sliding window technique. It is possible that different approaches might lead to lower values. Similarly, higher dose rate for IMRT would linearly reduce the beam on time but not the effective treatment time, which is dominated by the beam multiplicity and the dead times necessary to move from one field to the next.
IMRT
RA1
RA2
MU/fraction Ratioa
1531 ± 206 1
468 ± 95 0.31 ± 0.07
545 ± 80 0.36 ± 0.07
DR mean [MU/min]
300
GS mean [°/s]
N/A
428.1 ± 90.4 [231.6, 600.0] 5.5 ± 0.2 [4.4, 5.5]
252.9 ± 72.9 [89.7, 538.1] 5.5 ± 0.0 [5.5, 5.5]
MU/° mean
N/A
1.3 ± 0.3 [0.7, 2.3]
0.8 ± 0.2 [0.3, 1.6]
Conclusions
Delivery time [min]
9.4 ± 1.7
1.1 ± 0.0
2.6 ± 0.0b
RapidArc was investigated for anal canal cancer in comparison to IMRT. Although excellent dosimetric results were achieved with IMRT, RapidArc with double arc allowed to add improvement in sparing some organs at risk and uncompromised target coverage.
a
With respect to MU/fraction of IMRT. b Incorporates 25 s fixed time to re-configure linac for second arc, including couch and collimator rotation to start position.
124
Volumetric modulater arc theraphy for anal canal cancer
Single arcs resulted to be slightly inferior, on target coverage and on sparing some OARs. Double arcs are acceptable also in terms of treatment time since, although longer with respect to single arc, it results 3.5 times shorter than average treatment time for IMRT. Clinical investigations should be performed to assess the actual benefit in terms of reduced toxicity or improved control shown in the physical planning studies. Disclosure Dr. L. Cozzi acts as Scientific Advisor to Varian Medical Systems and as Head of Research and Technological Development to Oncology Institute of Southern Switzerland, IOSI, Bellinzona. Acknowledgement The current investigation was partially covered by a Varian research grant to IOSI. References [1] Otto K. Volumetric modulated arc therapy: IMRT in a single arc. Med Phys 2008;35:310–7. [2] Cameron C. Sweeping-window arc therapy: an implementation of rotational IMRT with automatic beam-weight calculation. Phys Med Biol 2005;50:4317–36. [3] Cotrutz C, Kappas C, Webb S. Intensity modulated arc therapy (IMAT) with centrally blocked rotational fields. Phys Med Biol 2000;45:2185–206. [4] 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. [5] 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. [6] 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. [7] Duthoy W, De Gersem W, Vergote K, Coghe M, Botenberf T, De Deene Y, et al. Whole abdominal radiotherapy (WAPRT) using intensity modulated arc therapy (IMAT): first clinical experience. Int J Radiat Oncol Biol Phys 2003;57:1019–32. [8] Yu CX. Intensity-modulated arc therapy with dynamic multileaf collimation: an alternative to tomotherapy. Phys Med Biol 1995;40:1435–49. [9] 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.
[10] 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. [11] 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;89:254–62. [12] 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 three-dimensional conformal radiotherapy. Int J Radiat Oncol Biol Phys 2008;72:996–1001. [13] Chen YJ, Liu A, Tsai P, Vora N, Pezner R, Shultheiss T, et al. Organ sparing by conformal avoidance intensity modulated radiation therapy for anal cancer: dosimetric evaluation of coverage of pelvis and inguinal/femoral nodes. Int J Radiat Oncol Biol Phys 2005;63:274–81. [14] Milano M, Jani A, Farrey K, Rash C, Heimann R, Chmura S. Intensity modulated radiation therapy in the treatment of anal cancer: toxicity and clinical outcome. Int J Radiat Oncol Biol Phys 2005;63:354–61. [15] Menkarios C, Azria D, Laliberte B, Llacer Moscardo C, Gourgou S, Lemanski C, et al. Optimal organ sparing intensity modulated radiation therapy regimen for the treatment of locally advanced anal canal carcinoma: a comparison of conventional and IMRT plans. Radiat Oncol 2007;2:41. [16] 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. [17] Bosset J, Roelofsen F, Morgan D, Budach V, Coucke P, Jager J, et al. Shortened irradiation scheme, continuous infusion of 5-fluorouracil and fractionation of mitomycin C in locally advanced anal carcinomas. Results of a phase II study of the European organization for Research and Treatment of Cancer. Radiotherapy and Gastrointestinal Cooperative Groups. Eur J Cancer 2003;39:45–51. [18] 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. [19] 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. [20] 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. [21] 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. [22] 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. [23] Convery D, Rosenbloom M. The generation of intensity-modulated fields for conformal radiotherapy by dynamic collimation. Phys Med Biol 1992;37:1359–74. [24] Spirou S, Chui C. A gradient inverse planning algorithm with dose-volume constrains. Med Phys 1998;25:321–33.