BJR
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
British Journal of Radiology (2004) 77, 377-386
© 2004 British Institute of Radiology
doi: 10.1259/bjr/66104316
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mott, J H
Right arrow Articles by Logue, J P
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mott, J H
Right arrow Articles by Logue, J P

Full Paper

Development of a simultaneous boost IMRT class solution for a hypofractionated prostate cancer protocol

J H Mott, MSc1, J E Livsey, BM, BCh, FRCR2 and J P Logue, MBChB, FRCR2

1 North Western Medical Physics and 2 Clinical Oncology, Christie Hospital NHS Trust, Manchester M20 4BX, UK


   Abstract
Top
 Abstract
Introduction
 Materials and method
 Results
 Discussion
 Conclusion
 References
 
The purpose of this work was to develop a robust technique for planning intensity-modulated radiation therapy (IMRT) for prostate cancer patients who are to be entered into a proposed hypofractionated dose escalation study. In this study the dose escalation will be restricted to the prostate alone, which may be regarded as a concurrent boost volume within the overall planning target volume (PTV). The dose to the prostate itself is to be delivered in 3 Gy fractions, and for this phase of the study the total prostate dose will be 57 Gy in 19 fractions, with 50 Gy prescribed to the rest of the PTV. If acute toxicity results are acceptable, the next phase will escalate doses to 60 Gy in 20 x 3 Gy fractions. There will be 30 patients in each arm. This work describes the class solution which was developed to create IMRT plans for this study, and which enabled the same set of inverse planning parameters to be used during optimization for every patient with minimal planner intervention. The resulting dose distributions were compared with those that would be achieved from a 3D conformal radiotherapy (3DCRT) technique that used a multileaf collimator (MLC) but no intensity modulation to treat the PTV, followed by a sequential boost to raise the prostate to 57 Gy. The two methods were tested on anatomical data sets for a series of 10 patients who would have been eligible for this study, and the techniques were compared in terms of doses to the target volumes and the organs at risk. The IMRT method resulted in much greater sparing of the rectum and bladder than the 3DCRT technique, whilst still delivering acceptable doses to the target volumes. In particular, the volume of rectum receiving the minimum PTV dose of 47.5 Gy was reduced from a mean value of 36.9% (range 23.4% to 61.0%) to 18.6% (10.3% to 29.0%). In conclusion, it was found possible to use a class solution approach to produce IMRT dose escalated plans. This IMRT technique has since been implemented clinically for patients enrolled in the hypofractionated dose escalation study.


   Introduction
Top
 Abstract
 Introduction
Materials and method
 Results
 Discussion
 Conclusion
 References
 
Several retrospective studies [1, 2] have shown that control rates for prostate cancer may be improved by escalating dose in patients at intermediate-to-high risk. The limiting factor for dose escalation is normal tissue toxicity, with the rectum being the major organ at risk (OAR), particularly when the whole planning target volume (PTV) receives the escalated dose [3, 4]. A prospective trial by Pollack et al [5] showed an increase in 5 year biochemical no evidence of disease (bNED) of 27% when escalating doses for the whole PTV from 70 Gy to 78 Gy for patients with prostate specific antigen (PSA)>10 ng ml–1. In comparison, workers using a 2-phase technique to cone down the boost fields found this produced a more modest increase in the incidence of late effects [6].

The use of intensity-modulated radiotherapy (IMRT) to produce a simultaneous boost technique offers the possibility of escalating doses whilst simultaneously shaping dose around the rectum in order to limit toxicity. Zelefsky et al [7] used IMRT to treat nearly 800 patients to doses in excess of 80 Gy. They reported that both acute and late rectal toxicities were significantly reduced compared with 3D conformal radiotherapy (3DCRT) treatments at similar dose levels, whilst short-term PSA control rates were at least comparable. They concluded that, "based on the favourable risk:benefit ratio", IMRT should be the standard method of treatment delivery for localized prostate cancer.

The increased doses reported in the published literature have mainly been achieved by the addition of 1.8–2 Gy fractions. For doses which can exceed 86 Gy, this has led to extended treatment times of up to 8 to 9 weeks [7]. Recent analyses of clinical results by Brenner and Hall [8, 9] and by Fowler et al [10] have indicated that the {alpha}/{beta} ratio for prostate cancer may be lower than was previously thought, ranging between 1.4 Gy and 1.9 Gy, although this is still the subject of much debate [11, 12]. The assumption of a low {alpha}/{beta} ratio would mean that prostate cancer behaves as a late reacting tissue. Such findings have renewed interest in using hypofractionated radiotherapy treatment. In addition to possible radiobiological gains, shorter timescales for treatment delivery and reduced numbers of delivered fractions would lead to markedly improved patient convenience and substantial savings in resources [13].

A hypofractionation trial by Kupelian et al [14, 15] compared 70 Gy in 28 fractions using IMRT against 78 Gy in 39 delivered by 3DCRT and demonstrated comparable preliminary biochemical relapse-free survival (bRFS) rates with reduced rectal toxicity in the IMRT arm. Although the vast majority of these patients were of stage T1-T2a, and were therefore not in the patient group generally considered likely to benefit most from dose escalation [16], this does suggest that IMRT may allow the ability to decrease normal tissue toxicity without compromising tumour control.

The standard prostate treatment at this institute is 50 Gy in 16 fractions to the entire PTV [13, 17]. Assuming an {alpha}/{beta} of 1.5 and negligible effects of overall time changes, this equates to 70.2 Gy in 1.8 Gy fractions, or to 66 Gy in 2 Gy fractions (standard dose in the UK at this time) to which it has been shown to give comparable results in terms of tumour control [18] and normal tissue toxicity [19]. More recently, patients at intermediate or high risk have been treated to a boost dose of 54 Gy in 18 fractions (54 Gy/18f) following neo-adjuvant hormones. These patients were planned using a sequential boost 3DCRT technique, similar to that used in the RT01 trial [20], in which 48 Gy/16f was delivered to the whole PTV, followed by a 6 Gy/2f boost to the prostate alone.

With this background a dose escalation study has been instituted aiming to boost prostate doses up to 60 Gy in 20 fractions (equivalent to 76 Gy in 2 Gy fractions if {alpha}/{beta} is taken as 1.5). The study has two arms, with doses to the prostate being increased in stages initially to 57 Gy/19f and then to 60 Gy/20f, while the rest of the PTV is prescribed a nominal dose of 50 Gy. Patients eligible for the dose escalation study are those with a medium to poor prognosis, T3, N0, M0, or T2, N0, M0 with either or both a Gleason score>7 and 20<PSA<50 ng ml–1. They must have a World Health Organization (WHO) performance status of 0 or 1 and will all receive neo-adjuvant hormone therapy for at least 3 months prior to radiotherapy to a maximum of 6 months in total. Toxicity will be assessed for 30 patients in the 57 Gy arm before proceeding to the next dose level. Once it has been confirmed that the hypofractionated, escalated doses can be delivered without increasing the rectal or bladder toxicity beyond acceptable limits, i.e. if there is <=1 instance of Grade 3 toxicity in each case, it is intended to compare the best arm of this hypofractionated study with a conventionally fractionated dose escalation schedule in a proper randomized clinical setting.

An initial investigation looking at the possibility of using a concurrent IMRT boost to treat both levels in the dose escalation study had suggested that it would be possible to create IMRT plans that allowed doses to the bulk of the rectum to remain constant even though the boost dose had increased [21]. This paper describes the development of an IMRT class solution which was designed specifically for this trial, and explains some of the reasoning behind the final choice of dose limits and penalties. Plans were created by means of a 5-field simultaneous boost technique. By using a simultaneous, or concurrent, boost method, the PTV and the clinical target volumes (CTVs) could be assigned different doses, which could be delivered during a single fraction, in this case using a multileaf collimator (MLC) based step and shoot delivery technique. For this study, 3DCRT plans giving a boost of 57 Gy in 19 fractions were also created for each patient data set, and the results compared with the new IMRT technique, with particular regard to the rectal doses that would be received.

Class solutions for IMRT
Standardized treatment planning procedures for standard or conformal planning methods make the planning process more efficient and encourage consistency between plans produced for individual patients and by different planning staff. By applying the same approach to the inverse planning process it may be possible to achieve an acceptable plan for each patient whilst avoiding the prohibitive logistical consequences of individually optimizing the planning parameters. A "class solution" can be defined as a set of IMRT planning parameters (beam arrangements, dose limits and penalties) that can be applied to every patient. Class solutions may be used as a starting point for individual optimization of the dose limits, although for the dose escalation protocol it was hoped to find a set of parameters that could be applied to all patients with minimum intervention on a patient-by-patient basis. The class solution therefore needed to be robust enough to produce a dose distribution within suitable clinical tolerances regardless of individual patient anatomy.


   Materials and method
Top
 Abstract
 Introduction
 Materials and method
Results
 Discussion
 Conclusion
 References
 
Data for 10 patients who had been treated sequentially with the 54 Gy boost was used in this investigation. All patients had been CT scanned using 5 mm slices and treated in a supine position. Both the IMRT and the 3DCRT plans were produced using the Plato 3D treatment planning system (Nucletron B.V., The Netherlands). For each patient, two CTVs were defined. CTV1 included both the prostate and the seminal vesicles, whilst CTV2, which was to be used for the boost, contained the prostate alone. In Figure 1Go the two CTVs are superposed, but they separate as the seminal vesicles appear in more superior CT slices. The PTV was grown from CTV1 using an expansion of 10 mm in all directions apart from posteriorly, where a 7 mm margin was used, following standard practice at this hospital. For both treatment techniques the aim was to treat the boost volume (CTV2) to a dose of 57 Gy in 19 fractions, while for the rest of the PTV the intended dose was 50 Gy (±5%). In reality, although the PTV should receive a minimum dose of at least 47.5 Gy, parts of the volume would have higher doses due to the unavoidable dose gradient from the edge of the PTV towards the higher dose boost volume.



View larger version (67K):
[in this window]
[in a new window]
  		Figure 1. Volumes used for planning prostate patients. At this level clinical target volume (CTV)1=CTV2, although these volumes separate as the seminal vesicles appear. The picture on the right illustrates how the overlap priorities govern which volume individual pixels are assigned to during the optimization of the intensity-modulated radiation therapy (IMRT) plans. PTV, planning target volume.

 
The OARs to be considered during the optimization were the rectum, bladder and femoral heads. The rectum was outlined from the anal verge to the rectosigmoid junction, and then subdivided into that part of the volume inside the PTV (RPTV), which must receive a target dose, and that outside the PTV, rectum2, which could be spared as much as possible. The whole bladder was also outlined, along with a separate volume, bladder2, which contained that part of the bladder not overlapping the PTV. These volumes are all illustrated in Figure 1Go.

Patients were selected on the basis of prostate cancer risk factors alone, with no exclusions for anatomical considerations, in order to ensure that the results would be as widely applicable as possible. This led to a wide variety of PTV shapes and volumes, with markedly differing degrees of PTV shaping around the rectum. The ranges and average sizes of some of the volumes outlined are given in Table 1Go. As expected, there was a strong correlation between the sizes of CTV1, CTV2 and the PTV in individual patients. However, no correlation was seen between the size of the PTV and the size of the rectum or bladder, or in the percentages of either of these volumes overlapping the PTV (RPTV and BPTV, respectively).


View this table:
[in this window]
[in a new window]
  		Table 1. Target and organ at risk volume parameters — average volumes and ranges for the 10 patients in this study. RPTV and BPTV are the percentage volumes of rectum and bladder, respectively, inside the planning target volume (PTV)

 
IMRT plans
The class solution for the dose escalation study reported here was developed by systematically varying the beam arrangements, dose limits, constraints and penalties used by the optimization algorithm to create the IMRT plans. The optimization algorithm in the Plato-ITP inverse planning module works by iteratively changing the radiation fluence within each beam to minimize a cost function based on the difference between the prescribed and calculated doses at each voxel in the patient. The contribution of each voxel to the total cost function can be influenced by a series of penalty factors, with separate factors assigned to the maximum and minimum doses for target organs. Only a maximum dose need be specified for OARs, although dose volume constraints (DVC) can also be included. An overlap priority table must also be specified, as voxels falling within more than one organ (for example bladder and PTV) can only be assigned to a single volume during the optimization procedure.

In each patient, the aim was to treat the entire PTV to at least 47.5 Gy, whilst sparing the rectum overlap from higher doses, in particular from doses approaching the level of the prostate boost. The prostate should receive doses within ±5% of the prescribed boost dose (57 Gy). Three-, five-, seven- and nine-field beam arrangements, equally spaced with either a direct posterior or a direct anterior field, were explored. Three-field plans did not give the possibility of shaping within the high dose volume (to reduce the dose received by the rectum overlap) and 7 or 9 beam arrangements were not found to offer any significant advantage over 5 beams, and were thought to be less efficient in terms of the time which would be needed to deliver and verify the plans. As Stein et al [22] predicted, beam arrangements containing a direct posterior beam were found to spare the rectum to a greater degree than arrangements containing an anterior field, whilst anterior arrangements reduced the dose received by the bladder. As the rectum was considered to be the primary OAR, a posterior arrangement was selected for the class solution. After extensive investigations, it was found that a five-field arrangement with beams at 35°, 105°, 180°, 255° and 325° produced acceptable dose distributions in all patients. The beams were almost equally spaced although the lateral posterior beams were rotated slightly to avoid the treatment couch.

The volumes, overlap priorities, dose limits, penalties and dose volume constraints used in the class solution for the 57 Gy boost cohort are shown in Table 2Go. Previous investigations had shown that it was sometimes useful to create a further volume, rectum+, when creating IMRT plans, which consisted of the rectum expanded using a 10 mm margin in all directions. This was found to be particularly helpful when there was a large amount of PTV shaping around a small rectum, as it prevented high doses wrapping around behind the OAR.


View this table:
[in this window]
[in a new window]
  		Table 2. Optimization parameters for the simultaneous boost intensity-modulated radiation therapy (IMRT) class solution. During optimization, only those parts of each volume not overlapped by an organ with a higher priority (1 being the highest priority) are considered, as shown in Figure 1Go. For example, only that part of clinical target volume (CTV)1 not also inside CTV2, i.e. the seminal vesicles, will be assigned CTV1 doses and penalties, as CTV2 (the prostate) has a higher priority. Volumes marked as organ type T are treated as target organs during optimization, those marked O are organs at risk

 
The dose limits used in the optimization were much more demanding than those which would be accepted clinically. For example, a maximum dose of 59.9 Gy and a minimum dose of 54.2 Gy (±5% of 57 Gy) to the CTV2 would generally be considered acceptable. Tighter dose limits were considered to be necessary for two reasons. First, as there were conflicting constraints applied to the adjacent OARs, it was sometimes not possible, even with high penalty values, to achieve the desired target doses and some compromise was required. Second, the optimization algorithm does not attempt any further improvement to the doses within a given volume once that dose is within the limits specified. This particular optimization algorithm produces the theoretical fluence for each beam, which must then be "sequenced" into a series of deliverable segments. This last step may result in a small degree of plan degradation when the plans are recalculated using the deliverable fluence maps. Therefore, when setting out the optimization parameters, it was found to be helpful to ask for tighter tolerances than those actually required, so that this would still result in a satisfactory dose distribution once the above limitations had been taken into account.

Our protocol dictated that the remainder of the CTV1 and PTV (those parts not also in the CTV2 boost volume) should be prescribed a nominal dose of 50 Gy, indicating that a clinically acceptable minimum dose would be 47.5 Gy. However, as these volumes were both adjacent to the prostate, the maximum dose they received would necessarily be the lowest dose received by the boost volume. Therefore, the maximum allowed dose for these volumes was increased to 54 Gy, i.e. to just below the minimum dose that would be acceptable in CTV2. Without relaxation of this constraint the planning system attempted to "sharpen" the dose gradient between the PTV or CTV1 and CTV2, resulting in highly modulated fluence maps which could only be delivered using many small segments. Likewise, the RPTV volume maximum dose needed to be increased to 54 Gy as this volume was also immediately adjacent to the prostate. An additional dose volume constraint of 50%<50 Gy was added to RPTV in order to keep doses as low as possible, consistent with achieving a minimum dose of at least 47.5 Gy. Since the RPTV is actually part of the PTV, it was initially included in the target organ list with a specified minimum dose. However, investigations showed that this tended to result in increased doses to rectum2 (the part of the rectum not in the PTV) and that treating the RPTV as an OAR, with a very high priority on the maximum dose, resulted in optimum placement of the 95% isodose, which consistently conformed tightly to the posterior edge of the PTV.

Rectum2, rectum+ and bladder2 were all specified a maximum dose of 95% of the PTV dose, as ideally they should be outside the treated volume [23]. To help further minimize rectal doses, rectum2 was also assigned a dose limit of 50%<45 Gy, which is that traditionally applied to the whole rectum at this hospital when creating standard 4-field 3DCRT plans.

The femoral heads were assigned a maximum dose of 30 Gy (chosen to be comparable with doses received from standard dose 3DCRT plans) but a very low penalty weighting. The body, in this case everything else not outlined already, had a maximum dose of 50 Gy. In general, the dose contribution to the target was fairly evenly distributed between the five beams, so setting the body dose limit lower in order to discourage subcutaneous hotspots was not found to be necessary. If the maximum dose limit to the body had been reduced it may then have been necessary to differentiate between the body voxels immediately adjacent to the PTV and the rest of the body remote from the high dose region.

When plans are created using IMRT it is not appropriate to prescribe to the isocentre, as the dose here will not necessarily be representative of the mean target dose and could reasonably be anywhere within the acceptable homogeneity limits of the target dose. Prescribing to the mean target dose was found to be more representative of the actual dose range received. Therefore, in the case of the simultaneous boost IMRT plans, plans were optimized to achieve a mean CTV2 dose of 57 Gy ±1%. The aim of each plan was to treat the PTV to at least 47.5 Gy (95% of 50 Gy) and the CTV2 to between 54.2 Gy and 59.9 Gy (±5% of 57 Gy).

It is interesting to note that our class solution gives the OARs equal or higher importance than dose uniformity within the target organs. This approach leads to a very robust solution for this combination of disease site, planning system and delivery method.

3DCRT plans
3DCRT plans were created by means of a 2-phase sequential boost technique. Using anterior, posterior and bilateral MLC fields, 48 Gy/16f was delivered to the whole PTV, followed by a 9 Gy boost using the same field arrangement but with field sizes coned down to cover the prostate alone. 48 Gy is slightly less than the desired PTV dose of 50 Gy but was chosen partly to fit in with the 3 Gy fraction schedule and also to compensate for the additional dose to the adjacent OARs from the boost fields. For both phases, the MLC leaves were fitted using a penumbra margin of 6 mm in all directions apart from superiorly and inferiorly, where it was increased to 10 mm to account for the shallower dose gradient in these directions. For each phase, doses were prescribed to an isocentre placed at the centre of the relevant target volume, and treatments were planned to cover the target volume with ±5% of the prescribed dose. When the two phases were combined, this meant that CTV2 would be covered by 95% of the total dose (57 Gy) and the PTV by 95% of the initial 48 Gy plus a contribution from the boost fields.

Plan comparison
An example of the dose distributions produced for a typical patient data set is given in Figure 2Go, which shows doses for the IMRT simultaneous boost technique and the two phases of the 3DCRT sequential boost technique. The plans were assessed in terms of the doses received by the target organs and the OARs, using isodoses, dose statistics and dose volume histograms. Maximum and minimum doses were recorded as those received by 1% of the organ volume.



View larger version (94K):
[in this window]
[in a new window]
  		Figure 2. Dose distributions for a central slice of a typical patient: (a) the intensity-modulated radiation therapy (IMRT) simultaneous boost technique—isodoses represent 95% of the prescribed planning target volume (PTV) dose (50 Gy) and 95% of the prostate boost dose (57 Gy); (b) phase 1 of the 3D conformal radiotherapy (3DCRT) sequential boost technique (isodose 95% of 48 Gy PTV dose); (c) phase 2 of the 3DCRT technique (isodose 95% of 9 Gy boost volume dose). In all cases the contours outlined are the clinical target volume (CTV) (at this level CTV1=CTV2), PTV, bladder (bl) and rectum (rec).

 

   Results
Top
 Abstract
 Introduction
 Materials and method
 Results
Discussion
 Conclusion
 References
 
Target doses
Table 3Go shows the maximum, mean and minimum doses in the target volumes achieved using the two techniques. For the IMRT plans, the average mean CTV2 dose was 56.9 Gy with good consistency in the maximum and minimum doses to CTV2 between the 10 patients. The average maximum dose for the patients was 58.6 Gy, with all patients falling within the +5% limit (59.9 Gy). The average minimum dose to CTV2 was 54.1 Gy. This falls slightly below the –5% limit of 54.2 Gy, but is marginally above the maximum allowed dose of 54 Gy to RPTV, which was assigned a much higher importance weighting than the minimum dose to the prostate during optimization. The lowest minimum CTV2 dose in any patient was 53.3 Gy. The minimum doses to CTV1 were maintained at a raised level due to the proximity of the seminal vesicles to the boost region and the relaxation of the maximum dose constraint for the CTV1 during the optimization process.


View this table:
[in this window]
[in a new window]
  		Table 3. Target dose statistics (in Gy) for the intensity-modulated radiation therapy (IMRT) and 3D conformal radiotherapy (3DCRT) plans. The average of the 10 patients is given with the range in parenthesis. 95% of the planning target volume (PTV) dose is 47.5 Gy, 95% of the clinical target volume (CTV)2 boost dose is 54.2 Gy. An asterisk (*) indicates those parameters in which a statistically significant difference (p<0.05) was found between the two techniques, using the Wilcoxon matched pairs signed rank test

 
There was more variation in the mean dose to the PTV between patients as, depending on the size and shape of the prostate, seminal vesicles and PTV, a variable portion of this volume was prescribed the boost dose. However, there was still good consistency in the minimum dose received. The average minimum dose of 46.8 Gy was marginally lower than the traditional limit of –5%, which would result in a dose of 47.5 Gy, but it is generally accepted that the benefits of increased OAR sparing through the use of IMRT can only be achieved at the expense of greater inhomogeneity within the target volume. Examination of the IMRT dose distributions showed that the voxels receiving the lowest doses were always those furthest away from the prostate and therefore least likely to contain disease. In effect, there was a steep dose gradient across the PTV towards the high dose boost region. Action to obtain greater coverage of the PTV would reduce the rectal sparing achieved, and it was observed that the IMRT plans still resulted in higher minimum PTV doses than the alternative 3DCRT approach.

For the sequential boost 3DCRT technique, the average mean dose to the prostate (CTV2) was 57.0 Gy and both the maximum and minimum doses were within ±5% in every patient. There was more variation in the doses received by CTV1 as this volume also included the seminal vesicles, which were prescribed a dose of 48 Gy. The minimum dose of 48.1 Gy to CTV1 was markedly less than that for the IMRT plans but was still high when compared with that which would be clinically acceptable due to the close proximity of the seminal vesicles to the boost volume. The minimum dose to the PTV was only 46.4 Gy. In principle, this could have been as low as 45.6 Gy (95% of 48 Gy), but most of the PTV would also receive a dose contribution from the 9 Gy boost fields. Lower PTV doses were particularly noticeable in patients with large seminal vesicles, in whom the superior portion of the PTV was at some distance from the boost fields. Increasing the dose delivered by the initial MLC fields would have increased the minimum dose to the PTV but would also have increased both rectal and bladder doses, so a compromise was required.

The isocentres for the two phases of the 3DCRT technique were typically very close together, therefore the dose at the CTV2 isocentre was very close to 57 Gy in every patient. The average mean CTV2 dose for the 3DCRT plans was also 57.0 Gy, varying between patients from 56.5 Gy to 57.5 Gy. In comparison, the simultaneous boost IMRT plans were prescribed to the average CTV2 dose (allowing a tolerance of ±1%). Although this resulted in a mean CTV2 dose close to 57 Gy in all patients (see Table 3Go), the isocentre doses varied quite widely, ranging from 56.0 Gy to 58.2 Gy, and averaging 57.3 Gy. This variation in the isocentre dose between patients was the reason for choosing different prescription practices for the two techniques.

Organs at risk
The percentage volumes of rectum and rectum2 irradiated to particular doses for each of the boost techniques are shown in Figure 3Go. In order to reduce rectal doses in the IMRT boost plans, during optimization a DVC of 50%<45 Gy was applied to rectum2, with a DVC of 50%<50 Gy applied to RPTV. This strategy resulted in nine of the 10 IMRT patient plans achieving the standard clinical objective of 50%<45 Gy for the whole rectum, with, on average, only 34.8% (range 22.0–54.7%) of the rectum receiving 45 Gy or more. The average rectal volume receiving 95% of the prescribed PTV dose (V47.5) was 18.6% (10.3–29.0%). The exact figure for each patient depended to some extent on the percentage of rectum inside the PTV volume, but, as shown in Table 1Go, for this series of patients the average RPTV was 16.6%, which correlates well with V47.5. Furthermore, on average only 4.5% (1.3–15.2%) of rectum2 received the minimum PTV dose, and hence the application of this class solution could be seen to produce good shaping of the high doses around the posterior edge of the PTV, resulting in maximal rectal sparing. None of the patients had any volume of rectum2 receiving doses approaching the boost level. More importantly, the volumes of the whole rectum receiving >95% of the boost volume dose were extremely low, averaging 0.9% (0.1–1.6%), despite the close proximity of the rectum to the prostate and that the average minimum dose in the boost volume was 54.1 Gy. This meant that for each patient plan there was good placement of the 95% boost isodose along the junction between the prostate and the rectum as well as excellent shaping of the 95% PTV isodose. This shaping of the dose within the treated volume to spare the rectum overlap could be seen in the isodose distributions of all patients, and was one of the main objectives for the design of the class solution.



View larger version (12K):
[in this window]
[in a new window]
  		Figure 3. Percentage volumes of (a) the rectum and (b) rectum2 irradiated to 45.0 Gy, 47.5 Gy (95% of the planning target volume nominal dose) and 54.2 Gy (95% of the prostate boost dose) for the intensity-modulated radiation therapy (IMRT) and 3D conformal radiotherapy (3DCRT) techniques.

 
In contrast, the irradiated volumes for the 3DCRT technique were much higher. For the whole rectum, the average volume receiving 45 Gy or more was relatively high (43.3%, range 27.0–67.5%), and three patients did not meet the 50%<45 Gy constraint. The average V47.5 for the whole rectum was 36.9% (range 23.4–61.0%), almost twice that seen when using the IMRT technique. Of greater concern was the fact that, on average, a considerable 8.8% (range 4.9–13.8%) of the rectal volume received more than 95% of the boost dose (54.2 Gy). For rectum2 the relevant volumes were 32.0%>45 Gy (range 17.1–58.6%), 25.0%>47.5 Gy (13.0–50.6%) and 0.6%>54.2 Gy (0.0–2.1%).

The average percentage volumes of the bladder irradiated to 47.5 Gy were 41.9% (range 21.4–64.2%) and 58.5% (27.0–79.8%) for the IMRT and 3DCRT techniques, respectively, whilst the average percentage volume for the bladder/PTV overlap was 36.8% of the bladder volume. Again, there was good correlation on a patient-to-patient basis between the amount of bladder inside the PTV and V47.5 for the IMRT technique.

The average maximum dose to the femoral heads for the IMRT technique was 27.3 Gy (22.6 Gy to 32.9 Gy). The majority of patients had maximum doses within the 30 Gy limit set during the optimization, with only one patient having both femoral heads exceeding this maximum and three further patients having one femoral head receiving more than 30 Gy. Furthermore, the maximum femoral head dose received in any patient was only 32.9 Gy, even though the dose limit was given a very low importance weighting in the class solution. In contrast, doses received using the 3DCRT technique were comparatively high, with an average maximum dose of 35.8 Gy (range 33.0–38.8 Gy). The femoral head doses would be expected to be higher for the 3DCRT technique partly because they were irradiated by two out of four beams, as opposed to two out of five beams for the IMRT arrangement.

Figure 4Go shows dose–volume histograms (DVH) for a typical patient, which illustrate the differences between the IMRT simultaneous boost method (solid lines) and the 3DCRT sequential boost MLC technique (dashed lines). For both plans there is reasonable homogeneity within the prostate boost volume (CTV2), although the 3DCRT plan has a slightly more uniform dose (Figure 4aGo). As discussed previously, the minimum dose to the prostate for the IMRT plan lay just outside the –5% limits usually recommended, but this was an unavoidable effect of the greater rectal sparing seen with the IMRT plan. The DVHs for both plans show a small amount of under-dosing of the PTV, although the IMRT method generated a smaller volume outside the 47.5 Gy limit. The low dose region of the PTV in the 3DCRT plan is an inevitable consequence of the sequential boost technique, as some areas would only be covered by the MLC fields delivering 48 Gy in 16 fractions of the base dose, and would receive minimal dose from the boost fields. Other parts of the PTV which are irradiated with the boost MLC fields but outside the high dose volume receive an increased dose. In contrast, the smooth dose gradient seen from the edge of the PTV up to the CTV2 boost volume with the IMRT method is reflected in the shape of the DVH curve.



View larger version (18K):
[in this window]
[in a new window]
  		Figure 4. Dose–volume histograms (DVHs) for a typical prostate patient. Solid lines indicate the 57 Gy intensity-modulated radiation therapy (IMRT) technique, dashed lines the 3D conformal radiotherapy (3DCRT) technique. DVHs are shown for (a) the prostate (clinical target volume (CTV)2) and planning target volume (PTV); (b) bladder (bl); (c) rectum (rec) and (d) left femoral head (fhl).

 
The DVHs for the rectum and bladder for the 3DCRT technique show the step effect typical of 4-field anterior–posterior and bilateral beam directions (Figures 4b,cGo). DVHs for these organs with standard 4-field MLC plans show a characteristic discontinuity at half the target dose, as part of their volume will be inside the PTV and will therefore receive target doses, while the remainder will be outside the target volume but still irradiated with one of the parallel opposed pair of beams. This step can still be seen in the boost plans here, although the dose at which the step appears is increased, as most of the OARs were also irradiated by the anterior and posterior boost MLC fields. The DVHs for the IMRT plan show a different, much smoother shape, due in part to the different beam orientation employed. The largest difference between the plans was that the volumes irradiated to doses above 47.5 Gy were greatly reduced with the IMRT rather than the 3DCRT technique, due to the improved shaping of the isodoses around the rectum, following the contour of the PTV. Although a small percentage of the rectum must receive at least the minimum dose given to the prostate, this volume was reduced by the use of dose limits and DVCs applied to the overlap region and the rest of the rectum. These DVHs are typical of patients planned using the class solution. The same general shape was seen on all patients, although the exact volumes depended on individual patient anatomy, particularly the amount of rectum inside the PTV.


   Discussion
Top
 Abstract
 Introduction
 Materials and method
 Results
 Discussion
Conclusion
 References
 
The additional shaping in the patient's cross-section afforded by IMRT meant that only that part of the rectum immediately adjacent to the prostate needed to receive the minimum prostate dose. This effect had previously been demonstrated by Amer et al [21] who, in modelling rectal movement incorporating positional uncertainties for a typical patient, calculated that the normal tissue complication probability (NTCP) (>=grade 2 RTOG [24]) would remain constant at around 5% for the initial 54 Gy boost and both levels of the dose escalation study if using an IMRT technique. It was claimed that the smaller dose per fraction received by the bulk of the rectum balanced out the effect of the higher doses received by a very small volume as the prostate boost dose increased. In contrast, a 3DCRT method like that described above produced NTCPs of 6.6%, 8.4% and 10.6% for the three escalated dose levels. Although this work predicted that the 3DCRT technique would achieve higher tumour control probabilities (TCPs), (with increases of 12%, 24% and 37% for the three dose levels compared with 5%, 16% and 22% for the IMRT technique), the higher NTCPs indicated that the conformal technique may not have been clinically acceptable for the boost levels in the dose escalation study.

It may be argued that some of the improvement in rectal sparing seen with the IMRT technique could be attributed to the change in beam number and directions compared with the four-field 3DCRT technique. However, using the same five fields as the IMRT technique but applying the sequential conformal boost method would in fact increase the amount of rectum irradiated to high doses, due to the subsequent change in the overall shape of the isodoses. For example, for the patient whose DVHs were given in Figure 4Go, the 5-field IMRT technique and the 4-field 3DCRT technique would irradiate 23.7% and 47.7% of the rectum to 47.5 Gy, respectively, whereas for a 5-field 3DCRT technique using the same beam directions as the IMRT method, the amount of rectum irradiated to this level would increase to 58.2%. It is the isodose shaping that is possible with IMRT, rather than the change in beam directions, which gives the IMRT technique the advantage in reducing rectal volumes irradiated to high doses.

For the sequential MLC boost technique, the entire PTV was planned to receive only 48 Gy ±5% in 16 fractions, therefore some parts of the PTV remote from the additional boost fields received minimum doses slightly lower than the clinical aim of 47.5 Gy. In contrast, the minimum dose received by the PTV in the IMRT method, although marginally higher, would be delivered over the full 19 fractions, as this technique used a simultaneous boost. However, inspection of the isodose distributions for each patient confirmed that the minimum doses to the PTV with the IMRT technique were always at the edges of the PTV, particularly where it had been expanded lateroposteriorly to cover the seminal vesicles. In effect, there was a continuous dose gradient across the PTV margin from its outer edge towards the prostate boost volume. This meant that the part of the PTV which was nearest to the prostate (and therefore most likely to contain disease) would get a higher dose than those areas of the PTV remote to the boost volume.

This paper describes an IMRT class solution which was developed to make planning large numbers of patients a realistic proposition. As inverse planning does not necessarily result in the doses originally asked for, it is important to consider conflicts and tolerances when setting up the optimization parameters. Despite the wide variety in organ volumes and shapes between patients, the IMRT class solution developed for this study produced acceptable plans in all patients with excellent consistency in the maximum and minimum doses to target organs between patients. This is very encouraging for the use of a class solution approach where using the same inverse planning parameters for every patient dramatically reduces the time needed to plan individual patients. The average number of total segments used in the IMRT plans created for this study was 64 (range 56–76). The average number of monitor units (mu) required for the IMRT plans was 866 mu per fraction, compared with 448 mu for the 3DCRT plans, although it should be noted that this could be substantially increased by the inclusion of conventional wedges in the 3DCRT plans.


   Conclusion
Top
 Abstract
 Introduction
 Materials and method
 Results
 Discussion
 Conclusion
References
 
It has been demonstrated that an IMRT class solution can provide treatment plans meeting clinical requirements for dose escalation using a hypofractionated concurrent boost technique. The advantage of such an approach is that an acceptable plan can be produced in a reasonable time frame and ensures that all patients are treated in a very similar way, making their comparison more straightforward. In particular, doses to critical structures can be constrained within the required tolerances without compromising target coverage. Both the IMRT and 3DCRT techniques were found to produce acceptable doses to the CTV2 boost region and the remainder of the PTV, but the minimum dose to the prostate was slightly lower using the IMRT technique, consistent with the greater rectal sparing seen. The volume of rectum receiving 95% of the original PTV dose was almost halved by the isodose shaping facilitated by the IMRT technique, and less than 1% of the rectum received 95% of the boost dose (compared with 8.8% using the 3DCRT method). Doses to the bladder and the femoral heads were also substantially reduced.

The simultaneous boost IMRT technique has now been introduced clinically at our institute for patients in the 57 Gy arm of the dose escalation study. The class solution described in this paper has been used as a starting point for every patient and in practice has been found to be robust, requiring minimal fine-tuning on a patient-by-patient basis. It has been possible for all patients recruited to be planned this way.


   Acknowledgments
 
The authors would like to thank Mr J M Wilkinson for his help in the preparation of this manuscript.


   Footnotes
 
Current address for Ms J H Mott, Regional Medical Physics Department, Newcastle General Hospital, Newcastle-upon-Tyne NE4 6BE, UK. Back

Received for publication November 29, 2002. Revision received June 18, 2003. Accepted for publication October 3, 2003.


   References
Top
 Abstract
 Introduction
 Materials and method
 Results
 Discussion
 Conclusion
 References
 

  1. Hanks GE, Hanlon AL, Schultheiss TE, Pinover WH, Movsas B, Epstein BE, et al. Dose escalation with 3D conformal treatment: five-year outcomes, treatment optimisation, and future directions. Int J Radiat Oncol Biol Phys 1998;41:501–10.[CrossRef][Medline]
  2. Zelefsky MJ, Leibel SA, Gaudin PB, Kutcher GJ, Fleschner NE, Venkatramen ES, et al. Dose escalation with three-dimensional conformal radition therapy affects the outcome in prostate cancer. Int J Radiat Oncol Biol Phys 1998;41:9491–500.
  3. Pollack A, Zagars GK, Starkschall G, Antolak JA, Lee JJ, Huang E, et al. Prostate cancer radiation dose response: results of the M. D. Anderson Phase III randomised trial. Int J Radiat Oncol Biol Phys 2002;53:1097–105.[CrossRef][Medline]
  4. Jackson A, Skwarchuk MW, Zelefsky MJ, Cowen DM, Venkatramen ES, Levegrun S, et al. Late rectal bleeding after conformal radiotherapy of prostate cancer (II): volume effects and dose-volume histograms. Int J Radiat Oncol Biol Phys 2001;49:685–98.[CrossRef][Medline]
  5. Pollack A, Zagars GK, Smith LG, Lee JJ, von Eschenbach AC, Antolak JA, et al. Preliminary results of a randomised radiotherapy dose escalation study comparing 70 Gy with 78 Gy for prostate cancer. J Clin Oncol 2000;18:3904–11.[Abstract/Free Full Text]
  6. Michalski J, Purdy J, Winter K, Roach M 3rd, Vijayakumar S, Sandler HM, et al. Preliminary report of toxicity following 3D radiation therapy for prostate cancer on 3DOG/RTOG 9406. Int J Radiat Oncol Biol Phys 2000;46:391–402.[CrossRef][Medline]
  7. Zelefsky MJ, Fuks Z, Hunt M, Yamada Y, Marion C, Ling CC, et al. High-dose intensity modulated radiation therapy for prostate cancer: early toxicity and biochemical outcome in 772 patients. Int J Radiat Oncol Biol Phys 2002;53:1111–6.[CrossRef][Medline]
  8. Brenner DJ, Hall EJ. Fractionation and protraction for radiotherapy of prostate carcinoma. Int J Radiat Oncol Biol Phys 1999;43:1095–101.[CrossRef][Medline]
  9. Brenner DJ. Towards optimal external-beam fractionation for prostate cancer. Int J Radiat Oncol Biol Phys 2000;48:315–6.[Medline]
  10. Fowler J, Chappell R, Ritter M. Is alpha/beta for prostate tumours really low? Int J Radiat Oncol Biol Phys 2001;50:1021–31.[CrossRef][Medline]
  11. D'Souza WD, Thames HD. Is the {alpha}/{beta} ratio for prostate cancer low? Int J Radiat Oncol Biol Phys 2001;51:1–3.
  12. Wang JZ, Guerrero M, Li XA. How low is the alpha/beta ratio for prostate cancer? Int J Radiat Oncol Biol Phys 2003;55:194–203.[CrossRef][Medline]
  13. Logue JP, Cowan RA, Hendry JH. Hypofractionation for prostate cancer. Int J Radiat Oncol Biol Phys 2001;49:1522–3.[Medline]
  14. Kupelian PA, Reddy CA, Klein EA, Willoughby TR. Short-course intensity-modulated radiotherapy (70 Gy at 2.5 Gy/fraction) for localized prostate cancer: preliminary results on late toxicity and quality of life. Int J Radiat Oncol Biol Phys 2001;51:988–93.[Medline]
  15. Kupelian PA, Reddy CA, Carlson TP, Altsman KA, Willoughby TR. Preliminary observations on biochemical relapse-free survival rates after short-course intensity-modulated radiotherapy (70 Gy at 2.5 Gy/fraction) for localized prostate cancer. Int J Radiat Oncol Biol Phys 2002;53:904–12.[CrossRef][Medline]
  16. Hurwitz MD, Schnieder L, Manola J, Beard CJ, Kaplan ID, D'Amico AV, et al. Lack of radiation dose response for patients with low-risk clinically localized prostate cancer: a retrospective analysis. Int J Radiat Oncol Biol Phys 2002;53:1106–10.[Medline]
  17. Read G, Pointon RC. Retrospective study of radiotherapy in early carcinoma of prostate. Br J Urol 1989;63:191–5.[Medline]
  18. Logue JP, Livsey JE, Cowan RA, Wiley JP, Swindell R, Read G, et al. Hypofractionated conformal radiotherapy in carcinoma of the prostate: 5 year outcome analysis. Int J Radiat Oncol Biol Phys 2002;54:S189.
  19. Livsey JE, Routledge J, Burns M, Swindell R, Cowan RA, Logue JP, et al. Scoring of treatment related late effects in prostate cancer. Radiother Oncol 2002;65:109–21.[CrossRef][Medline]
  20. MRC (Medical Research Council). RT01: a randomised trial of high dose therapy in localised cancer of the prostate using conformal radiotherapy techniques. MRC radiotherapy working party, London. 1998.
  21. Amer AM, Mott J, Mackay RI, Williams PC, Hendry JH, Livsey J, et al. Prediction of the benefits from dose-escalated hypofractionated intensity-modulated radiotherapy for prostate cancer. Int J Radiat Oncol Biol Phys 2002;56:199–207.[CrossRef]
  22. Stein J, Mohan R, Wang XH, Bortfeld T, Wu Q, Preiser K, et al. Number and orientations of beams in intensity-modulated radiation treatments. Med Phys 1997;24:149–60.[Medline]
  23. ICRU (International Commission on radiation Units, Measurements). Prescribing, Recording and Reporting Photon Beam Therapy (Supplement to ICRU Report 50). ICRU Report 62. 1999.
  24. Cox JD, Stetz J, Pajak TF. Toxicity criteria of the Radiation Therapy Oncology Group (RTOG) and the European Organization for Research and Treatment of Cancer (EORTC). Int J Radiat Oncol Biol Phys 1995;31:1341–6.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 Br. J. Radiol.Home page
R M WYATT, B J JONES, and R G DALE
Radiotherapy treatment delays and their influence on tumour control achieved by various fractionation schedules
Br. J. Radiol., July 1, 2008; 81(967): 549 - 563.
[Abstract] [Full Text] [PDF]

Home page
 Br. J. Radiol.Home page
BJR Review of the Year - 2004
Br. J. Radiol., March 1, 2005; 78(927): 181 - 185.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mott, J H
Right arrow Articles by Logue, J P
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mott, J H
Right arrow Articles by Logue, J P

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
BJR DMFR IMAGING  ALL BIR JOURNALS