British Journal of Radiology (2005) 78, 358-361
© 2005 British Institute of Radiology
doi: 10.1259/bjr/56095273
Re-treatment of a lung tumour using a simple intensity-modulated radiotherapy approach
A W Beavis, PhD
1
A Abdel-Hamid, FRCR
2 and
S Upadhyay, MD
2
1 Department of Medical Physics, Hull and East Yorkshire NHS Trust and Clinical Biosciences Institute, University of Hull, Princess Royal Hospital, Saltshouse Road, Kingston Upon Hull HU8 9HE and 2 Department of Clinical Oncology, Princess Royal Hospital, Saltshouse Road, Kingston Upon Hull HU8 9HE, UK
 |
Abstract
|
|---|
We present a lung tumour case where, although the proximal spinal cord had already reached its dose tolerance, re-treatment was indicated. Minimization of the cord dose was defined as the main constraint for this case, however additional dose to the heart was considered. A simple (4 field) intensity-modulated radiotherapy (IMRT) treatment which proved superior to the standard conformal plan was developed using the Computerized Medical Systems (CMS) XiO treatment planning system. The IMRT plan was found to be superior to the conventional conformal plan regarding tumour coverage. It provided 100% saturation of the planning target volume (PTV) by the 95% isodose cloud, whereas the latter only provided 77% coverage. A step and shoot delivery using 10 intensity levels was developed and subsequently delivered for this patient. We considered it to be a routine application of IMRT and an important example of how it can offer benefit in individual and appropriate cases where conventional treatment is inadequate.
|
Case study
|
|---|
Recently, a patient presented with moderately to poorly differentiated squamous carcinoma recurrent lung tumour. He had responded to initial radiotherapy, however he was symptomatic with disease progression and further treatment was indicated to improve his quality of life; he declined chemotherapy. The previous treatment meant the cord had already received its dose tolerance, so any further radiotherapy required that it received "no further dose". The heart had also received the full prescribed dose. Various plans were considered, including simple two-dimensional (2D) and conformal three-dimensional (3D) plans, it became clear that the only option that could provide the homogeneous dose required to the planning target volume (PTV) whilst simultaneously avoiding giving unacceptable doses to the cord and heart was an intensity-modulated radiotherapy (IMRT) plan.
Treatment planning options
At our Cancer Centre all computerized treatment planning is performed using the XiO (previously FOCUS) treatment planning system from Computerized Medical Systems (St. Louis, MO). This system has an integrated IMRT option/inverse planning module, which essentially comprises an optimization algorithm, a leaf sequencing algorithm and quality assurance (QA) options [1]. These component parts enable the design of intensity modulated beams that satisfy a required dose be delivered to both the PTV and dose is restricted to the proximal organs at risk.
For this patient the rationale in developing the plan for this re-treatment course was simply to give "zero" dose to the spinal cord. The dose tolerance for the cord is considered to be 45 Gy when given in 2 Gy fractions, the dose it was given in the initial treatments is biologically equivalent to this. Furthermore, as a secondary consideration it was deemed desirable to minimize the dose to the heart and contralateral lung.
Initially, simple conformal options were attempted. The only way to achieve the cord goal was to ensure the organ was well shielded and no entrance or exit doses were incident to it. A two field conformal plan was developed that achieved these constraints, however the compromise was significant lack of coverage of the PTV. This trade off was not considered clinically acceptable.
An IMRT plan, following a conformal avoidance strategy, was developed using inverse planning. Four beam directions were chosen, the field portal defining collimators were manually set to totally shield the cord. Figure 1
shows the beam's eye view of these fields. Two of the beams did not irradiate the whole of the PTV, the inverse planning software was used to develop intensity modulated beams that essentially compensated for this, using the other two beams to provide a reasonable coverage through the PTV by "filling in the gaps".
View larger version (66K):
[in this window]
[in a new window]
|
Figure 1. The rectangular outlines in each beam's eye view denotes the field defining collimators. We use a gantry angle convention where a 0 degree beam direction defines an anterior field for a supine patient and angles increase in a clockwise direction about the patients left side. We note that each beam is forced to conformally avoid the (rendered) cord. The non-uniform outline denotes the maximum irradiated area under each of the intensity modulated beams.
|
|
The demand placed on the plan was to cover the PTV with the 95% isodose surface and to keep the volume irradiated to greater than the prescription dose as small as possible, idealistically we aim to have 0% of the PTV receiving more than 107%. A 6 MV beam set was used. Since the cord was not directly irradiated no specific dose constraints for it were defined in the optimizer. The doses in all other (non-target) tissues were controlled by a single constraint parameter in the cost-function; a maximum dose of 25 Gy was defined. This constraint was assigned a relative importance of 5% (as compared with that of the target), however the interplay between the terms in the cost function [2] is complicated and outside the scope of this manuscript.
The resulting beams were produced using a "Bortfeld-Boyer" [3] based leaf sequencing algorithm using 10 intensity levels and 0.5 cm "steps", segments smaller than 1 cm2 were automatically deleted before the final forward computation. The spatial resolution in the leaf travel direction is chosen to reflect the leaf widths of the multileaf collimator (MLC) used in our centre.
At this stage dose–volume histograms (DVH) are generated. The dose–volume profiles given to the PTV and organs at risk are assessed. Furthermore, the 3D dose distribution is consulted to determine the position of hot spots to assess their significance. Until an acceptable plan was obtained this entire optimization-through-DVH generation was iterated upon.
The treatment was delivered using a 600CD Clinac equipped with a Millennium 120 MLC and an aSi Electronic Portal Imaging, this equipment was manufactured by Varian Oncology Systems (Palo Alto, CA).
The "best" conventional plan, regarding cord avoidance, was not acceptable owing to spatially incomplete irradiation of the PTV. This plan utilized a wedged pair of posterior oblique fields at gantry angles of 132° and 200°. By employing the MLC to shield the cord in the beam's eye view a "zero" cord dose was obtained, however only 77% of the PTV received 28.5 Gy (95% of the prescribed dose to the isocentre) with the sub-volume proximal to the cord, receiving only approximately 13.3 Gy.
The IMRT plan that was accepted delivered 28.5 Gy to 100% of the PTV. The four field plan used fields at 70°, 90°, 140° and 180°. The maximum dose inside the PTV of 34.3 Gy was generated, a small 118% hot spot was evident lying just outside and posteriorly to the target volume. These were considered acceptable due to the absolute value of the prescription dose. The DVH curves generated for both plans are shown in Figure 2. The dose–volume profiles are displayed for the cord, heart and PTV. We note that similar distributions were achieved by either approach for the organs at risk. However, the target coverage given by the conventional treatment option is clearly inferior to that offered by the IMRT plan. The curve for the conventional treatment PTV implies that re-normalization of that plan was not a serious option in any attempt to improve the high dose coverage of the PTV. The lung doses from the two plans were comparable, the V20 for the IMRT plan was 1% and 2% for the two field conformal plan. This result was not surprising given the prescription dose was 30 Gy, to aid comparison with other techniques we computed a V20 of 2% having re-normalized the IMRT plan prescription to 55 Gy and 3% for 64 Gy.
View larger version (55K):
[in this window]
[in a new window]
|
Figure 2. Dose–volume histogram showing the dose–volume profiles for the planning target volume (PTV) and organs at risk for plans: (1) the intensity-modulated radiotherapy (IMRT) solution chosen and (2) the best conventional option regarding dose to the cord.
|
|
Figure 3 demonstrates the geometry of the problem posed by this patient's treatment. The dose cloud that obscures the PTV is that of the 95% isodose. The highest dose inside the "posterior protuberance" of the 90% isodose cloud was 32 Gy.
|
Discussion
|
|---|
In Figure 4 we show digital images, from our Odelft Simulator, obtained during an assessment of the patients suitability for conformal treatment. The upper row were anterior–posterior images, the lower row lateral images, those in the left column were obtained under normal respiration. The patient was asked to breathe deeply, then hold at deep inhalation (centre column) and again at deep expiration (right column). We note from the images that the cord–isocentre distances remain fairly constant. We did not consider tumour/organ motion a significant issue in the delivery of this treatment.
View larger version (79K):
[in this window]
[in a new window]
|
Figure 4. Digitally captured images from the simulator. The upper row were anterior–posterior (AP) images and the lower row lateral (LAT) images. The left column were obtained under normal respiration. The centre column on breath-hold at deep inhalation and the right column similarly at deep expiration.
|
|
Further to this proactive analysis, we should mention that our practice is to contour the cord canal and so a protective margin of several millimetres is always assigned to our cord volumes.
The choice of dose calculation algorithms is an important consideration [4, 5] when using "high end" planning methodologies and comparing one method with another, especially so in lung treatments. We used a convolution algorithm using the FFT (fast Fourier transform) approximations [6] that require invariant dose deposition kernels, as with all our current IMRT work. There is a very strong argument that a Full, or superposition, convolution algorithm [7, 8] should be used in order to give a better approximation of the dose delivered to the lung and to lung–proximal tissue interfaces. In our department we have a guiding philosophy governing the introduction of new services, wherein we do not change too much at once. Hence, rather than introduce an algorithm not yet routinely used we elected to "stick with" the FFT-convolution algorithm.
There are a variety of studies and methods reported [9–13] using IMRT for the initial treatment of lung tumours. Our report is concerned with a different clinical situation, that being a re-treatment following initial radiotherapy where cord tolerance was reached.
A final consideration and point worth discussing in the context of our approach is the option offered by forward planned [14] multiple segmented treatments. In some cases we choose to manually add segments to "fill in cold areas" in dose distributions. These cases are usually large volumes that cover most of the thoracic region. We considered this option when planning this patient. However, having developed a good deal of experience and confidence with inverse planned IMRT we concluded that it would be quicker and more efficient to have the optimization routines deal with the problem. This is not to say that a forward planned, approach is not valid, we simply feel that our solution was a better use of resources in our clinic.
|
Acknowledgments
|
|---|
We would like to thank Computerized Medical Systems for their invaluable support, practical and financial, and enthusiasm for our research and clinical development programmes. It is also a pleasure to acknowledge the support of the Clinical BioSciences Institute of the University of Hull and that of Yorkshire Cancer Research.
Received for publication September 12, 2003.
Revision received December 10, 2004.
Accepted for publication January 5, 2005.
|
References
|
|---|
- Webb S. Intensity modulated radiotherapy. Bristol: IOP Publishing, 2000.
- Xing L, et al. Fast iterative algorithms for 3D inverse treatment planning. Med Phys 1998;25:1845–9.[Medline]
- Bortfeld T, Boyer A. Realization and verification of three-dimensional conformation radiotherapy with modulated fields. Int J Radiat Oncol Biol Phys 1994;30:899–908.[Medline]
- Miften M, Beavis A, Marks L. The influence of the dose calculation model on treatment plan evaluation in conformal therapy: a three case study. Med Dosim 2002;27:51–7.[Medline]
- Jeraj R, Keall P, Siebers J. The effect of dose calculation accuracy on inverse treatment planning. Phys Med Biol 2002;47:391–407.[CrossRef][Medline]
- Boyer A, et al. Fast Fourier transform convolution calculations of x-ray isodose distributions in homogeneous media. Med Phys 1989;16:248–53.[Medline]
- Mackie T, Scrimger J, Battista J. A convolution method of calculating dose for 15 MV X-rays. Med Phys 1985;12:188–96.[CrossRef][Medline]
- Miften M, et al. Implementation of FFT convolution and multigrid superposition models in the FOCUS RTP system. Phys Med Biol 2000;45:817–33.[Medline]
- De Jaeger K, et al. Incorporating an improved dose-calculation algorithm in conformal radiotherapy of lung cancer: re-evaluation of dose in normal lung tissue. Radiother Oncol 2003;69:1–10.[Medline]
- Derycke S, et al. Non-coplanar beam intensity modulation allows large dose escalation in Stage III lung cancer. Radiother Oncol 1997;45:253–61.[Medline]
- Derycke S, et al. Conformal radiotherapy of Stage III non-small cell lung cancer: a class solution involving non-coplanar intensity modulated beams. Int J Radiat Oncol Biol Phys 1998;41:771–7.[CrossRef][Medline]
- Dirkx M, Heijmen BJ. Beam intensity modulation for penumbra enhancement and field length reduction in lung cancer treatments: a dosimetric study. Radiother Oncol 2000;56:181–8.[Medline]
- Zaider M, Amols HI. A little to a lot or a lot to a little: is NTCP always minimized in multiport therapy? Int J Radiat Oncol Biol Phys 1998;41:945–50.[CrossRef][Medline]
- Xia P, Pickett B, Vigneault E, Verhey LJ, Roach M 3rd. Forward or inversely planned segmental multileaf collimator IMRT and sequential tomotherapy to treat multiple dominant intraprostatic lesions of prostate cancer to 90 Gy. Int J Radiat Oncol Biol Phys 2001;51:244–54.[Medline]
Top
Abstract
Case study
