- Email: [email protected]
Treatment margins and treatment fractionation in conformal radiotherapy of muscle-invading urinary bladder cancer
Radiotherapy and Oncology 71 (2004) 65–71 www.elsevier.com/locate/radonline
Treatment margins and treatment fractionation in conformal radiotherapy of muscle-invading urinary bladder cancer Ludvig Paul Murena,b,*, Anthony Thomas Redpathb, Duncan Bruce McLarenc a
Section of Oncology, Institute of Medicine, Medical Faculty, University of Bergen, and Department of Oncology and Medical Physics, Haukeland University Hospital, Bergen, Norway b Department of Oncology Physics, Directorate of Clinical Oncology, Western General Hospital, Edinburgh, UK c Department of Radiation Oncology, Directorate of Clinical Oncology, Western General Hospital, Edinburgh, UK Received 13 May 2003; accepted 10 November 2003
Abstract Background and purpose: Different treatment margins and fractionation schedules are used in conformal radiotherapy (CRT) of urinary bladder cancer. This study compared intestine and rectum dose – volume histogram (DVH) data and normal tissue complication probability (NTCP) estimates for various clinically applied margins and fractionation schedules in bladder irradiation. Patients and methods: Normal tissue dose distributions in fifteen bladder cancer patients treated with CRT were studied using standard three- and four-field configurations. The impact of margin width on intestine and rectum dose distributions was initially evaluated using DVH data. NTCP modelling with the probit model was used to compare the impact of choice of margin size and fractionation schedule. The analysis included margin combinations of 1.0 cm isotropic (narrow margins) and 1.2 –2.0 cm non-isotropic (wide margins) and fractionation schedule alternatives of 52.5 Gy/20, 55 Gy/20, 57.5 Gy/20 and 64 Gy/32. Results: Using wide as compared to narrow margins, the volumes of intestine and rectum receiving high doses increased by factors of approximately two and four, respectively. Similar differences between wide and narrow margins were found when calculating intestine and rectum NTCPs. The impact of margin size depended strongly on the volume effect expressed by the NTCP model parameters. With standard parameters, however, the choice of margins and fractionation schedule had a similar impact on intestine NTCPs, while for the rectum, the choice of margin had a greater impact than the choice of fractionation. For a given margin size, the intestine and rectum NTCPs for the 55 Gy/20 and the 64 Gy/32 schedules were comparable. For clinics using narrow margins and a fractionation of 52.5 Gy/20, the NTCP modelling suggested that a change in fractionation schedule (to 55 Gy/20 or 64 Gy/32) or a change to wide margins would have a similar effect on the intestine NTCP predictions. Conclusions: This modelling study documented that the choice of margins was as important as the choice of fractionation in terms of intestine and rectum DVH data and NTCP predictions. q 2004 Elsevier Ireland Ltd. All rights reserved. Keywords: Urinary bladder cancer; Conformal radiotherapy; Normal tissue doses and complication probabilities; Treatment margins; Fractionation schedules
1. Introduction The optimal management of muscle-invading urinary bladder cancer, i.e. primary radical surgery versus radical radiotherapy (RT), remains an area of great debate [5,29]. In addition, if conformal radiotherapy (CRT) is prescribed, there is a large variation worldwide in the size of the treatment margins and the daily doses used [16,24]. Narrow treatment margins (typically 1 cm isotropic) are often used in combination with a large daily dose and short treatment * Corresponding author. 0167-8140/$ - see front matter q 2004 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.radonc.2004.02.003
times (52.5–57.5 Gy in 20 fractions, i.e., approx. 2.6–2.9 Gy per fraction), an approach widespread in the UK [3,8]. Elsewhere, RT of bladder cancer is prescribed with conventional fractionation (60–70 Gy in 30–35 fractions, 2 Gy per fraction) and often with wider margins (1.5–2.0 cm) to account better for the documented geometrical uncertainties (internal bladder and external patient motion) [17,19]. Comparisons between the tumour-sterilising effect of RT with various treatment margins and fractionation schedules may be confounded by differences in patient selection between the clinical series, i.e., radical RT used as a primary vs. secondary treatment option. In this modelling
66
L.P. Muren et al. / Radiotherapy and Oncology 71 (2004) 65–71
study, attention is instead given to the degree of normal tissue irradiation when using different margin and fractionation alternatives. Traditionally, normal tissue dose–volume histograms (DVHs) have been used to evaluate treatment plans. Mathematical models of the radiation dose response of normal tissues have been presented to condense the DVH into an estimated normal tissue complication probability (NTCP) in the organ e.g. [12,13,15]. Despite the limitations of current NTCP models (e.g. sparse prospectively recorded follow-up data, lack of spatial information, variations in organ definitions [7,18,28]) they are believed to be helpful as supplementary tools for evaluation of RT. In this study, we therefore used one frequently applied NTCP model with parameters obtained by fitting to clinical data. Due to the many pitfalls in NTCP modelling, the NTCP estimates were correlated to the basic DVH data from which they were derived. The underlying view of centres using narrow margins and prescribing a large daily dose is that the narrow margins compensate for the potentially detrimental normal tissue effects of large fraction sizes. The aim of this study was therefore to compare the impact of choice of treatment margins and choice of treatment fractionation on intestine and rectum DVH data and NTCP predictions in standard CRT of bladder cancer.
2. Materials and methods 2.1. Selection of patients Fifteen patients with muscle-invading urinary bladder cancer (14 males, one female) referred consecutively to the Edinburgh Cancer Centre at Western General Hospital (WGH) for radical RT in the period April to October 2002 were included. Their age ranged from 48 to 82 years (mean 72 years). They were treated according to the standard threefield WGH CRT technique and prescribed with the standard WGH fractionation schedule (see below). 2.2. Definitions of target volumes and normal tissues All organs/structures were outlined in VirtSim, an in-house developed treatment planning system [22,23]. For the purpose of this study, the clinical target volume (CTV), i.e., the bladder, was accurately re-drawn by a radiation oncologist delegated responsibility for bladder irradiation (DMcL). Using a margin-growing feature of VirtSim, two sets of planning target volumes (PTVs) were derived from the CTVs. One set was constructed with CTV – PTV margins of 1.0 cm isotropic (narrow margins); the other with 1.2 –2.0 cm non-isotropic (1.2 cm laterally, 2.0 cm elsewhere; wide margins). The two sets were used to construct two treatment plan series, with narrow and wide margins, respectively. The CT planning studies acquired in the supine position consisted of 5 mm thick slices with 5 mm distance between
the centre of the slices, and covered all patients from well above the promontory to well below the caudal pelvic floor. All segments of the intestine were included up to the level of the promontory. For the rectum we used the first slice below the recto-sigmoid flexure as superior limit, and the first slice above the anal verge as inferior limit [2,7,11,14,19,25], and included the rectal wall and the contents of the rectum in the volume. All normal tissue structures were outlined by a radiotherapy physicist (LPM) and reviewed by the radiation oncologist (DMcL). 2.3. Planning of standard three-field and four-field treatment techniques Planning calculations of the standard CRT techniques of both WGH and Haukeland University Hospital (HUH), Bergen, Norway were performed in VirtSim. The three-field WGH technique consisted of anterior and wedged lateral beams, with beam weights of typically 70 and 40– 50, respectively, and typically 60-degree wedges. In the fourfield HUH technique a posterior field was added to the WGH configuration, the wedge angles were reduced to around 30 – 40 degrees, and beam weights were typically 40– 50, 30 –40 and 40 –50 for the anterior, posterior and lateral beams, respectively. Beam weights were specified at the reference point (maximum dose on the central axis) to achieve 100% at iso-centre. For both techniques the MLC configuration in each beam portal was set up automatically using a pre-defined margin to account for the penumbra effect (5 mm ant/post and lateral, 7 mm sup/inf). 2.4. Fractionation schedules used in bladder irradiation Using the two standard treatment techniques, combinations of the two margin alternatives and a range of clinically relevant fractionation schedules were evaluated and compared. The current standard WGH and HUH fractionations, 52.5 Gy/20 fractions and 64 Gy/32 fractions, respectively, were supplemented with two other hypo-fractionated schedules, 55 Gy/20 fractions and 57.5 Gy/20 fractions, that have been/still are in use in the UK. 2.5. Evaluation of treatment plans: DVH-analysis and NTCP modelling Initially, the volumes of intestine and rectum above various dose levels—most of them corresponding to published DVH constraints [20]—were used to compare the two margin alternatives. The intestine constraints were based on the studies of Gallagher et al. [10] and Baglan et al. [1], and included the following cut-off points: 30 Gy to 135 cm3; 45 Gy to 78 cm3 and 50 Gy to 17 cm3. Note that the intestine constraints were specified in absolute volumes, due to the large extent of the intestine. Using the actual fractionation schedules underlying these constraints, the dose levels in the constraints were a/b-corrected to apply to
L.P. Muren et al. / Radiotherapy and Oncology 71 (2004) 65–71
the 64 Gy/32 fractionation schedule (a/b ¼ 3 Gy). The resulting dose levels then corresponded to 48, 77 and 84% of the prescription dose at the 64 Gy/32 schedule. We therefore quantified the absolute intestine volumes above these three levels and above the prescription dose, i.e., VInt;48% ; VInt;77% ; VInt;84% and VInt;100% ; respectively. The rectum constraints were derived from a number of recent clinical studies, mostly on prostate irradiation, and were formulated as permitted fractional volumes above various dose levels [20]. The applied cut-off points were the following: 48 Gy to 70% of the rectum volume; 58 Gy to 55%; and 65.6 Gy to 31% [20]. After a/b-correction to the 64 Gy/32 fractionation, the resulting dose levels corresponded to 73, 88 and 100% of the prescription dose at the 64 Gy/32 schedule. To also quantify the fraction of rectum receiving lower doses we added the dose level corresponding to 30 Gy (47% of the prescription dose using the 64 Gy/ 32 schedule). For the rectum we therefore derived the fractional volumes receiving more than 47, 73, 88 and 100% of the prescription dose, i.e., vRect;47% ; vRect;73% ; vRect;88% and vRect;100% : The fulfilment of the DVH constraints for both organs was also scored. To compare the various combinations of margins and fractionation we performed NTCP calculations with the probit model [12,15] using procedures in VirtSim. The implemented probit model algorithm was confirmed against manual calculations using clinical DVHs. As the choice of treatment technique was found to have a minor impact on the resulting NTCPs, only the WGH technique was used in all of the NTCP calculations. The probit NTCP model was derived as a modification of the error function [15]. It was described with the following three parameters: D50 ; the uniform dose absorbed in the reference volume Vref causing a 50% probability of a complication; m; a parameter describing the slope of the dose response curve; and n; a volume dependency parameter. A high value of n indicated a strong volume effect in the specific tissue. The dose levels in the DVHs were corrected for fractionation (a/b ¼ 3 Gy) before the ‘effective volume’ reduction scheme proposed by Kutcher and Burman [12] was applied to obtain reduced, single-step histograms. For the intestine the probit model was used with parameters from Burman et al. (D50 ¼ 55 Gy; m ¼ 0:16; n ¼ 0:15) with Vref being the whole intestine [4,9]. Based on a previous repeat CT scanning study, the total intestine volume was estimated based on the axial cross-section of the patient at the level of the promontory [19]. A zero dose bin was added to the intestine DVH to get a total volume corresponding to the cross-section of the patient. For the rectum, we initially applied the probit model parameters from Burman et al. (D50 ¼ 80 Gy; m ¼ 0:15; n ¼ 0:12) with the whole rectum being the Vref [4,9]. Since recent reports have indicated both a stronger and a weaker volume effect in the rectum [2,6], the calculations were repeated with both n ¼ 0:06 and n ¼ 0:20:
67
3. Results 3.1. The impact of treatment margin The choice of treatment margin had a large impact on intestine and rectum DVH data for both treatment techniques (Table 1). Using wide as opposed to narrow margins, the absolute volumes of intestine receiving high doses ðVInt;48% 2 VInt;100% Þ increased by factors in the range 1.5 to 2.4. The fulfilment of intestine DVH constraints was also substantially reduced when using the widest margins, e.g., at the 77% dose level, 11 of 15 patients fulfilled the constraint when using narrow margins compared to only 5 when using wide margins. The width of the margins had in general a greater impact on DVH data for the rectum than for the intestine, apart from at the lowest dose level (Table 1). The fractional rectum volumes receiving the highest doses ðvRect;73% 2 vRect;100% Þ increased by factors in the range 3.6 to 5.0 when using wide rather than narrow margins. The rectum DVH constraints in the intermediate to high dose range were easily fulfilled, also when using the widest margins. A similar impact of the choice of margin was found for the intestine and rectum NTCPs with different fractionation alternatives (Table 2). In general, the intestine NTCPs were higher than the rectum NTCPs for the same margin and fractionation combination. When using the probit model with the Burman et al. data for the intestine, our calculations suggested that using wide as opposed to narrow margins increased the intestine NTCPs with factors in the range of 2.4 to 2.6 for the different fractionation alternatives. As in the DVH analysis, the impact of margin size on NTCPs was stronger for the rectum than for the intestine. Still the impact of margins depended strongly on the seriality parameter of the NTCP model. With wide Table 1 A comparison of intestine and rectum DVH-data using narrow and wide CTV –PTV margins applying standard 3- and 4-field treatment techniques. Average absolute intestine volumes and fractional rectum volumes above various relative dose levels, mostly corresponding to intestine and rectum DVH constraints, are shown. The fractions of patients fulfilling DVH constraints scaled to the 64 Gy/32 fractionation schedule are given in brackets Narrow margins
Wide margins
3-Field
4-Field
3-Field
4-Field
Intestine VInt;48% (cm3) VInt;77% (cm3) VInt;84% (cm3) VInt;100% (cm3)
85.5 (12/15) 61.3 (11/15) 56.6 (2/15) 31.1
102.4 (11/15) 61.5 (11/15) 56.6 (2/15) 27.7
132.7 (10/15) 108.0 (5/15) 102.9 (0/15) 69.2
153.4 (8/15) 108.2 (5/15) 103.5 (0/15) 67.2
Rectum VRect;47% (%) VRect;73% (%) VRect;88% (%) VRect;100% (%)
24.3 6.0 (15/15) 4.7 (15/15) 2.7 (15/15)
81.0 6.9 (15/15) 4.8 (15/15) 2.6 (15/15)
52.3 22.3 (15/15) 17.3 (15/15) 12.3 (14/15)
89.3 24.7 (15/15) 18.2 (15/15) 13.1 (14/15)
68
L.P. Muren et al. / Radiotherapy and Oncology 71 (2004) 65–71
Table 2 Mean and standard deviation (in brackets) of the intestine and rectum NTCPs (%) comparing treatment with combinations of narrow and wide CTV– PTV margins and various fractionation schedules with the 3-field technique 52.5 Gy/20 fxs
55 Gy/20 fxs
64 Gy/32 fxs
57.5 Gy/20 fxs
Narrow margins Intestine Probit model and Burman data
1.1 (1.3)
2.1 (2.4)
2.3 (2.6)
4.0 (4.3)
Rectum Probit model and Burman data ðn ¼ 0:12Þ Probit model and modified Burman data ðn ¼ 0:06Þ Probit model and modified Burman data ðn ¼ 0:20Þ
0.1 (0.2) 0.5 (0.7) 0.01 (0.04)
0.2 (0.4) 1.1 (1.5) 0.04 (0.07)
0.3 (0.4) 1.2 (1.6) 0.04 (0.07)
0.5 (0.9) 2.3 (3.0) 0.09 (0.2)
Wide margins Intestine Probit model and Burman data
2.9 (2.8)
5.3 (2.4)
5.8 (2.6)
9.5 (4.3)
Rectum Probit model and Burman data ðn ¼ 0:12Þ Probit model and modified Burman data ðn ¼ 0:06Þ Probit model and modified Burman data ðn ¼ 0:20Þ
0.8 (1.0) 1.9 (1.8) 0.3 (0.5)
1.6 (2.0) 3.9 (3.5) 0.6 (1.0)
1.8 (2.2) 4.2 (3.7) 0.7 (1.2)
3.2 (3.9) 7.5 (6.4) 1.3 (2.1)
margins the average NTCPs increased by a factor of 6 to 8 when using the standard Burman parameters, compared to factors in the ranges of 3 to 4 and 15 to 30 with modified seriality, n ¼ 0:06 and n ¼ 0:20; respectively. A strong correlation between individual DVH data and NTCP predictions was observed for intestine and rectum volumes above all dose levels assessed (Figs. 1 and 2). The volumes above the highest doses had a slightly stronger impact on the NTCP estimates (i.e., a larger gradient for the straight line fit) than the volumes above the lower dose level. 3.2. The impact of treatment fractionation The average NTCPs for both organs varied in general by factors in the range 3 to 6 across the different fractionation schedules (Table 2). A factor of approximately 9 was observed for one of the modified probit model parameterisations for the rectum ðn ¼ 0:20Þ; but this parameterisation resulted in very low average NTCP estimates. The 64 Gy/32 and the 55 Gy/20 schedules produced very similar NTCPs both for intestine and rectum. For clinics using narrow margins and a fractionation of 52.5 Gy/20, the NTCP modelling suggested that a change in fractionation schedule to 55 Gy/20 or 64 Gy/32 or a change to wide margin would have a similar effect on the intestine NTCP predictions.
4. Discussion This study compared intestine and rectum DVH data and NTCPs in bladder irradiation with a range of margins and fractionation schedule combinations. It was conducted using two widely used beam configurations, i.e. three-field
and four-field CRT arrangements, which led to very similar intestine and rectum irradiation (Table 1). As a consequence, the average NTCPs for the two techniques differed by less than 0.2% (probit model and standard Burman parameters, on all combinations of the 52.5 Gy/20 and the 64 Gy/32 schedules and narrow and wide margins; data not shown). On the other hand, the choice of margins and fractionation schedule was shown to cause large differences in the normal tissue doses/volumes. The choice of treatment margin increased the absolute volumes of intestine receiving a high dose by approximately a factor of two, while the fractional rectum volumes receiving a high dose increased even more, by approximately a factor of four (Table 1). This reflects the close proximity of the rectum to the treatment volume and the relatively small axial extent of the rectum. Still, scoring the fulfilment of a/b-corrected intestine and rectum DVH constraints (64 Gy/32 fractionation used) confirmed that the intestine is the major organ at risk in bladder irradiation, in line with clinical experience [24] and previous NTCP modelling [18]. The results of the NTCP analysis comparing the impact of variation in treatment margins (i.e. volume effects) and treatment fractionation (i.e. dose effects) depended heavily on the radiation sensitivity parameter of the NTCP model. Probably due to difficulties caused by the shape and mobility of the intestine, there has been few NTCP modelling studies for this organ. With the probit model we therefore only applied the standard Burman et al. data. Using these parameters, however, rather similar findings in relation to the impact of the size of margins were found in the NTCP modelling and the DVH analysis. The importance of the seriality parameter was clearly emphasised in the rectum NTCP analysis, as the calculations were repeated with three different n parameters, altering the volume effect/seriality. When using parameters expressing low
L.P. Muren et al. / Radiotherapy and Oncology 71 (2004) 65–71
69
Fig. 1. The correlation between four sets of intestine DVH-data from each of the 15 individual patients and their corresponding intestine NTCP estimate, using the probit model with the standard Burman parameters to evaluate the treatment data for the 64 Gy/32 fractionation schedule, wide margins and the 3-field technique. The linear regression curves were derived from the individual DVH-data for each of the dose levels. A strong correlation between DVH data and NTCP estimates was observed for the volumes above all dose levels. The volumes above the highest doses had a slightly stronger impact on the NTCP estimates (i.e., a larger gradient for the straight line fit) than the volumes above the lower dose level. The literature constraints at the intermediate dose levels are in agreement with the Burman parameters for the probit model, as they fall at an NTCP level of approximately 5%.
volume effect/high seriality ðn ¼ 0:06Þ; the variation in NTCPs between different margins was relatively low and similar to the variation between different fractionation schedules. However, for the standard Burman parameters and the parameterisation reflecting increased volume effect/lower seriality ðn ¼ 0:20Þ; the choice of margin caused considerably larger variation in the NTCPs than the choice of fractionation. Good agreement was observed also at the individual level between DVH and NTCP data (Figs. 1 and 2). Based on these relations, it is possible to estimate within 1% the predicted NTCP using one DVH parameter only. In addition, these relations can be used to derive the expected increase in NTCP after a change in the volume receiving at least a certain dose level. Intestine and rectum literature constraints were also included in Figs. 1 and 2. For the intestine, two of the constraints (at the 48 and 77% dose levels) were in agreement with the applied NTCP parameters and model since these constraints were found at an NTCP level of approximately 5%. The intestine constraint at the higher
dose level was very strict, both in the sense that it corresponded to a negligible NTCP level and in the sense that it was fulfilled in only a few of the cases analysed (Table 1). This might indicate that there is a need for updated, prospectively recorded intestine follow-up data, in particular for higher dose levels. For the rectum, all literature constraints were in agreement with the standard Burman parameters for the probit NTCP model, as they were located at an NTCP level of approximately 5%. When using the two modified parameterisations, the agreement between the NTCP model and the literature constraints was less good (data not shown). The NTCP data presented here did not allow any validation of the applied NTCP model. Nevertheless the average NTCPs appear to be in line with observed late adverse effect rates for both previous and current margin and fractionation combinations at WGH and HUH [8,26, 27]. For instance, the 57.5 Gy/20 schedule—though with standardised (typically 10 £ 10 cm2) field sizes—was abandoned due to a high intestinal adverse effect rate [8, 21]. Obviously, we advocate cautious interpretation of
70
L.P. Muren et al. / Radiotherapy and Oncology 71 (2004) 65–71
Fig. 2. The correlation between four sets of rectum DVH-data from each of the 15 individual patients and their corresponding rectum NTCP estimate, using the probit model with the standard Burman parameters (same treatment technique, margins and fractionation as in Fig. 1). The linear regression curves were derived from the individual DVH-data for each of the dose levels. A strong association between DVH data and NTCP estimates was observed also for the rectum, at least for the volumes above the three highest dose levels. All literature constraints were in agreement with the standard Burman parameters for the probit model, as they fall at an NTCP level of approximately 5%.
NTCP modelling. It still seems that clinics that irradiate bladder cancer using narrow margins and the 52.5 Gy/20 schedule may consider an increase in the margins to better account for geometrical uncertainties (bladder motion and patient positioning) or a moderate dose escalation. The current analysis suggests that the 55 Gy/20 and the 64 Gy/ 32 schedules are equivalent in terms of intestine and rectum irradiation. Calculations on a subset of the patients also confirmed that the 57.5 Gy/20 and a 68 Gy/34 schedule resulted in almost identical intestine and rectum NTCP estimates (data not shown). With the clinical experience of the 57.5 Gy/20 fractionation in mind, this also shows that escalation of the dose to the whole bladder above these levels should not be investigated without implementing highly conformal treatment techniques, such as intensity modulated RT (IMRT), that better spare the involved normal tissues. Treatment approaches such as IMRT and/or imageguided RT are currently being introduced to improved normal tissue sparing and if possible, to allow dose escalation. These techniques are likely to become of considerable benefit also in the management of bladder
cancer [20]. In addition to showing that these techniques probably are required for further dose escalation without increased morbidity in bladder irradiation, this study showed that large benefits can be realised by optimisation of more fundamental treatment parameters, such as the width of treatment margins and the choice of fractionation. This potential should be exploited before pursuing the more technologically sophisticated treatment approaches. In conclusion, this modelling study of bladder irradiation suggested that the choice of margins was as important as the choice of fractionation in terms of intestine and rectum DVH parameters and NTCP predictions. The 55 Gy/20 and 64 Gy/32 fractionation schedules appeared to be comparable in terms of intestine and rectum NTCP predictions.
Acknowledgements Generous grants from the Norwegian Cancer Society, The Norwegian foundation for Health and Rehabilitation, the University of Bergen and Haukeland University Hospital supported this study. The authors want to acknowledge the
L.P. Muren et al. / Radiotherapy and Oncology 71 (2004) 65–71
support from Olav Dahl, August Bakke, Jarle Rørvik and Ole Johan Halvorsen at Haukeland University Hospital and David Thwaites and Grahame Howard at Western General Hospital for the conduction of this study.
References [1] Baglan KL, Frazier RC, Yan D, et al. The dose–volume relationship of acute small bowel toxicity from concurrent 5-FU-based chemotherapy and radiation therapy for rectal cancer. Int J Radiat Oncol Biol Phys 2002;52:176–83. [2] Boersma LJ, van den Brink M, Bruce AM, et al. Estimation of the incidence of late bladder and rectum complications after high-dose (70–78 GY) conformal radiotherapy for prostate cancer, using dose– volume histograms. Int J Radiat Oncol Biol Phys 1998;41:83– 92. [3] Borgaonkar S, Jain A, Bollina P, et al. Radical radiotherapy and salvage cystectomy as the primary management of transitional cell carcinoma of the bladder. Results following the introduction of a CT planning technique. Clin Oncol (R Coll Radiol) 2002;14:141– 7. [4] Burman C, Kutcher G, Emami B, Goitein M. Fitting of normal tissue tolerance data to an analytic function. Int J Radiat Oncol Biol Phys 1991;21:123– 35. [5] Chahal R, Sundaram SK, Iddenden R, et al. A study of the morbidity, mortality and long-term survival following radical cystectomy and radical radiotherapy in the treatment of invasive bladder cancer in Yorkshire. Eur Urol 2003;43:246 –57. [6] Dale E, Hellebust TP, Skjønsberg A, Høgberg T, Olsen DR. Modeling normal tissue complication probability from repetitive CT scans during fractionated HDR brachytherapy and external beam radiotherapy of the uterine cervix. Int J Radiat Oncol Biol Phys 2000;47: 963–71. [7] Dale E, Olsen DR, Fossa˚ SD. Normal tissue complication probabilities correlated with late effects in the rectum after prostate conformal radiotherapy. Int J Radiat Oncol Biol Phys 1999;43:385–91. [8] Duncan W, Quilty PM. The results of a series of 963 patients with transitional cell carcinoma of the urinary bladder primarily treated by radical megavoltage X-ray therapy. Radiother Oncol 1986;7: 299–310. [9] Emami B, Lyman J, Brown A, et al. Tolerance of normal tissue to therapeutic irradiation. Int J Radiat Oncol Biol Phys 1991;21:109–22. [10] Gallagher MJ, Brereton HD, Rostock RA, et al. A prospective study of treatment techniques to minimize the volume of pelvic small bowel with reduction of acute and late effects associated with pelvic irradiation. Int J Radiat Oncol Biol Phys 1986;12:1565–73. [11] Jackson A, Skwarchuk MW, Zelefsky MJ, 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. [12] Kutcher GJ, Burman C. Calculation of complication probability factors for non-uniform normal tissue irradiation: the effective volume method. Int J Radiat Oncol Biol Phys 1989;16:1623–30.
71
˚ gren A, Brahme A. Tumor and normal tissue responses [13] Ka¨llman P, A to fractionated non-uniform dose delivery. Int J Radiat Biol 1992;62: 249–62. [14] Lebesque JV, Bruce AM, Kroes AP, et al. Variation in volumes, dosevolume histograms, and estimated normal tissue complication probabilities of rectum and bladder during conformal radiotherapy of T3 prostate cancer. Int J Radiat Oncol Biol Phys 1995;33:1109– 19. [15] Lyman JT. Complication probability as assessed from dose volume histograms. Radiat Res 1985;104:S13–S19. [16] Marks LB, Carroll PR, Dugan TC, Anscher MS. The response of the urinary bladder, urethra, and ureter to radiation and chemotherapy. Int J Radiat Oncol Biol Phys 1995;31:1257 –80. [17] Meijer GJ, Rasch C, Remeijer P, Lebesque JV. 3D analysis of delineation errors, setup errors and organ motion during radiation therapy of bladder cancer. Int J Radiat Oncol Biol Phys 2003;55: 1277– 87. [18] Muren LP, Jebsen N, Gustafsson A, Dahl O. Can dose response models predict reliable normal tissue complication probabilities in radical radiotherapy of urinary bladder cancer? The impact of alternative radiation tolerance models and parameters. Int J Radiat Oncol Biol Phys 2001;50:627– 37. [19] Muren LP, Smaaland R, Dahl O. Organ motion, set-up variation and treatment margins in radical radiotherapy of urinary bladder cancer. Radiother Oncol 2003;69(3):291 –304. [20] Muren LP, Smaaland R, Dahl O. Recent advances and future challenges in conformal radiotherapy of muscle-invading urinary bladder cancer. Radiother Oncol. Conditionally accepted. [21] Quilty PM, Duncan W, Kerr GR. Results of a randomised study to evaluate influence of dose on morbidity in radiotherapy for bladder cancer. Clin Radiol 1985;36:615–8. [22] Redpath AT. A beam model for three-dimensional radiotherapy planning. Br J Radiol 1995;68:1356–63. [23] Redpath AT. Planning of beam intensity modulation using an advanced 3D dose calculation algorithm and a simulated annealing method. Radiother Oncol 1998;49:295–304. [24] Sengelo¨v L, von der Maase H. Radiotherapy in bladder cancer. Radiother Oncol 1999;52:1–14. [25] Skwarchuk MW, Jackson A, Zelefsky MJ, et al. Late rectal toxicity after conformal radiotherapy of prostate cancer (I): multivariate analysis and dose-response. Int J Radiat Oncol Biol Phys 2000;47: 103–13. [26] Sma˚land R, Akslen LA, Tønder B, et al. Radical radiation treatment of invasive and locally advanced bladder carcinoma in elderly patients. Br J Urol 1991;67:61–9. [27] Tait DM, Nahum AE, Meyer LC, et al. Acute toxicity in pelvic radiotherapy; a randomised trial of conformal versus conventional treatment. Radiother Oncol 1997;42:121–36. [28] Ten Haken R. Introduction. Semin Radiat Oncol 2001;11:181– 3. [29] Zietman AL, Shipley WU. Organ-sparing treatment for bladder cancer: time to beat the drum. Int J Radiat Oncol Biol Phys 1994;30: 741–2.
Treatment margins and treatment fractionation in conformal radiotherapy of muscle-invading urinary bladder cancer Ludvig Paul Murena,b,*, Anthony Thomas Redpathb, Duncan Bruce McLarenc a
Section of Oncology, Institute of Medicine, Medical Faculty, University of Bergen, and Department of Oncology and Medical Physics, Haukeland University Hospital, Bergen, Norway b Department of Oncology Physics, Directorate of Clinical Oncology, Western General Hospital, Edinburgh, UK c Department of Radiation Oncology, Directorate of Clinical Oncology, Western General Hospital, Edinburgh, UK Received 13 May 2003; accepted 10 November 2003
Abstract Background and purpose: Different treatment margins and fractionation schedules are used in conformal radiotherapy (CRT) of urinary bladder cancer. This study compared intestine and rectum dose – volume histogram (DVH) data and normal tissue complication probability (NTCP) estimates for various clinically applied margins and fractionation schedules in bladder irradiation. Patients and methods: Normal tissue dose distributions in fifteen bladder cancer patients treated with CRT were studied using standard three- and four-field configurations. The impact of margin width on intestine and rectum dose distributions was initially evaluated using DVH data. NTCP modelling with the probit model was used to compare the impact of choice of margin size and fractionation schedule. The analysis included margin combinations of 1.0 cm isotropic (narrow margins) and 1.2 –2.0 cm non-isotropic (wide margins) and fractionation schedule alternatives of 52.5 Gy/20, 55 Gy/20, 57.5 Gy/20 and 64 Gy/32. Results: Using wide as compared to narrow margins, the volumes of intestine and rectum receiving high doses increased by factors of approximately two and four, respectively. Similar differences between wide and narrow margins were found when calculating intestine and rectum NTCPs. The impact of margin size depended strongly on the volume effect expressed by the NTCP model parameters. With standard parameters, however, the choice of margins and fractionation schedule had a similar impact on intestine NTCPs, while for the rectum, the choice of margin had a greater impact than the choice of fractionation. For a given margin size, the intestine and rectum NTCPs for the 55 Gy/20 and the 64 Gy/32 schedules were comparable. For clinics using narrow margins and a fractionation of 52.5 Gy/20, the NTCP modelling suggested that a change in fractionation schedule (to 55 Gy/20 or 64 Gy/32) or a change to wide margins would have a similar effect on the intestine NTCP predictions. Conclusions: This modelling study documented that the choice of margins was as important as the choice of fractionation in terms of intestine and rectum DVH data and NTCP predictions. q 2004 Elsevier Ireland Ltd. All rights reserved. Keywords: Urinary bladder cancer; Conformal radiotherapy; Normal tissue doses and complication probabilities; Treatment margins; Fractionation schedules
1. Introduction The optimal management of muscle-invading urinary bladder cancer, i.e. primary radical surgery versus radical radiotherapy (RT), remains an area of great debate [5,29]. In addition, if conformal radiotherapy (CRT) is prescribed, there is a large variation worldwide in the size of the treatment margins and the daily doses used [16,24]. Narrow treatment margins (typically 1 cm isotropic) are often used in combination with a large daily dose and short treatment * Corresponding author. 0167-8140/$ - see front matter q 2004 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.radonc.2004.02.003
times (52.5–57.5 Gy in 20 fractions, i.e., approx. 2.6–2.9 Gy per fraction), an approach widespread in the UK [3,8]. Elsewhere, RT of bladder cancer is prescribed with conventional fractionation (60–70 Gy in 30–35 fractions, 2 Gy per fraction) and often with wider margins (1.5–2.0 cm) to account better for the documented geometrical uncertainties (internal bladder and external patient motion) [17,19]. Comparisons between the tumour-sterilising effect of RT with various treatment margins and fractionation schedules may be confounded by differences in patient selection between the clinical series, i.e., radical RT used as a primary vs. secondary treatment option. In this modelling
66
L.P. Muren et al. / Radiotherapy and Oncology 71 (2004) 65–71
study, attention is instead given to the degree of normal tissue irradiation when using different margin and fractionation alternatives. Traditionally, normal tissue dose–volume histograms (DVHs) have been used to evaluate treatment plans. Mathematical models of the radiation dose response of normal tissues have been presented to condense the DVH into an estimated normal tissue complication probability (NTCP) in the organ e.g. [12,13,15]. Despite the limitations of current NTCP models (e.g. sparse prospectively recorded follow-up data, lack of spatial information, variations in organ definitions [7,18,28]) they are believed to be helpful as supplementary tools for evaluation of RT. In this study, we therefore used one frequently applied NTCP model with parameters obtained by fitting to clinical data. Due to the many pitfalls in NTCP modelling, the NTCP estimates were correlated to the basic DVH data from which they were derived. The underlying view of centres using narrow margins and prescribing a large daily dose is that the narrow margins compensate for the potentially detrimental normal tissue effects of large fraction sizes. The aim of this study was therefore to compare the impact of choice of treatment margins and choice of treatment fractionation on intestine and rectum DVH data and NTCP predictions in standard CRT of bladder cancer.
2. Materials and methods 2.1. Selection of patients Fifteen patients with muscle-invading urinary bladder cancer (14 males, one female) referred consecutively to the Edinburgh Cancer Centre at Western General Hospital (WGH) for radical RT in the period April to October 2002 were included. Their age ranged from 48 to 82 years (mean 72 years). They were treated according to the standard threefield WGH CRT technique and prescribed with the standard WGH fractionation schedule (see below). 2.2. Definitions of target volumes and normal tissues All organs/structures were outlined in VirtSim, an in-house developed treatment planning system [22,23]. For the purpose of this study, the clinical target volume (CTV), i.e., the bladder, was accurately re-drawn by a radiation oncologist delegated responsibility for bladder irradiation (DMcL). Using a margin-growing feature of VirtSim, two sets of planning target volumes (PTVs) were derived from the CTVs. One set was constructed with CTV – PTV margins of 1.0 cm isotropic (narrow margins); the other with 1.2 –2.0 cm non-isotropic (1.2 cm laterally, 2.0 cm elsewhere; wide margins). The two sets were used to construct two treatment plan series, with narrow and wide margins, respectively. The CT planning studies acquired in the supine position consisted of 5 mm thick slices with 5 mm distance between
the centre of the slices, and covered all patients from well above the promontory to well below the caudal pelvic floor. All segments of the intestine were included up to the level of the promontory. For the rectum we used the first slice below the recto-sigmoid flexure as superior limit, and the first slice above the anal verge as inferior limit [2,7,11,14,19,25], and included the rectal wall and the contents of the rectum in the volume. All normal tissue structures were outlined by a radiotherapy physicist (LPM) and reviewed by the radiation oncologist (DMcL). 2.3. Planning of standard three-field and four-field treatment techniques Planning calculations of the standard CRT techniques of both WGH and Haukeland University Hospital (HUH), Bergen, Norway were performed in VirtSim. The three-field WGH technique consisted of anterior and wedged lateral beams, with beam weights of typically 70 and 40– 50, respectively, and typically 60-degree wedges. In the fourfield HUH technique a posterior field was added to the WGH configuration, the wedge angles were reduced to around 30 – 40 degrees, and beam weights were typically 40– 50, 30 –40 and 40 –50 for the anterior, posterior and lateral beams, respectively. Beam weights were specified at the reference point (maximum dose on the central axis) to achieve 100% at iso-centre. For both techniques the MLC configuration in each beam portal was set up automatically using a pre-defined margin to account for the penumbra effect (5 mm ant/post and lateral, 7 mm sup/inf). 2.4. Fractionation schedules used in bladder irradiation Using the two standard treatment techniques, combinations of the two margin alternatives and a range of clinically relevant fractionation schedules were evaluated and compared. The current standard WGH and HUH fractionations, 52.5 Gy/20 fractions and 64 Gy/32 fractions, respectively, were supplemented with two other hypo-fractionated schedules, 55 Gy/20 fractions and 57.5 Gy/20 fractions, that have been/still are in use in the UK. 2.5. Evaluation of treatment plans: DVH-analysis and NTCP modelling Initially, the volumes of intestine and rectum above various dose levels—most of them corresponding to published DVH constraints [20]—were used to compare the two margin alternatives. The intestine constraints were based on the studies of Gallagher et al. [10] and Baglan et al. [1], and included the following cut-off points: 30 Gy to 135 cm3; 45 Gy to 78 cm3 and 50 Gy to 17 cm3. Note that the intestine constraints were specified in absolute volumes, due to the large extent of the intestine. Using the actual fractionation schedules underlying these constraints, the dose levels in the constraints were a/b-corrected to apply to
L.P. Muren et al. / Radiotherapy and Oncology 71 (2004) 65–71
the 64 Gy/32 fractionation schedule (a/b ¼ 3 Gy). The resulting dose levels then corresponded to 48, 77 and 84% of the prescription dose at the 64 Gy/32 schedule. We therefore quantified the absolute intestine volumes above these three levels and above the prescription dose, i.e., VInt;48% ; VInt;77% ; VInt;84% and VInt;100% ; respectively. The rectum constraints were derived from a number of recent clinical studies, mostly on prostate irradiation, and were formulated as permitted fractional volumes above various dose levels [20]. The applied cut-off points were the following: 48 Gy to 70% of the rectum volume; 58 Gy to 55%; and 65.6 Gy to 31% [20]. After a/b-correction to the 64 Gy/32 fractionation, the resulting dose levels corresponded to 73, 88 and 100% of the prescription dose at the 64 Gy/32 schedule. To also quantify the fraction of rectum receiving lower doses we added the dose level corresponding to 30 Gy (47% of the prescription dose using the 64 Gy/ 32 schedule). For the rectum we therefore derived the fractional volumes receiving more than 47, 73, 88 and 100% of the prescription dose, i.e., vRect;47% ; vRect;73% ; vRect;88% and vRect;100% : The fulfilment of the DVH constraints for both organs was also scored. To compare the various combinations of margins and fractionation we performed NTCP calculations with the probit model [12,15] using procedures in VirtSim. The implemented probit model algorithm was confirmed against manual calculations using clinical DVHs. As the choice of treatment technique was found to have a minor impact on the resulting NTCPs, only the WGH technique was used in all of the NTCP calculations. The probit NTCP model was derived as a modification of the error function [15]. It was described with the following three parameters: D50 ; the uniform dose absorbed in the reference volume Vref causing a 50% probability of a complication; m; a parameter describing the slope of the dose response curve; and n; a volume dependency parameter. A high value of n indicated a strong volume effect in the specific tissue. The dose levels in the DVHs were corrected for fractionation (a/b ¼ 3 Gy) before the ‘effective volume’ reduction scheme proposed by Kutcher and Burman [12] was applied to obtain reduced, single-step histograms. For the intestine the probit model was used with parameters from Burman et al. (D50 ¼ 55 Gy; m ¼ 0:16; n ¼ 0:15) with Vref being the whole intestine [4,9]. Based on a previous repeat CT scanning study, the total intestine volume was estimated based on the axial cross-section of the patient at the level of the promontory [19]. A zero dose bin was added to the intestine DVH to get a total volume corresponding to the cross-section of the patient. For the rectum, we initially applied the probit model parameters from Burman et al. (D50 ¼ 80 Gy; m ¼ 0:15; n ¼ 0:12) with the whole rectum being the Vref [4,9]. Since recent reports have indicated both a stronger and a weaker volume effect in the rectum [2,6], the calculations were repeated with both n ¼ 0:06 and n ¼ 0:20:
67
3. Results 3.1. The impact of treatment margin The choice of treatment margin had a large impact on intestine and rectum DVH data for both treatment techniques (Table 1). Using wide as opposed to narrow margins, the absolute volumes of intestine receiving high doses ðVInt;48% 2 VInt;100% Þ increased by factors in the range 1.5 to 2.4. The fulfilment of intestine DVH constraints was also substantially reduced when using the widest margins, e.g., at the 77% dose level, 11 of 15 patients fulfilled the constraint when using narrow margins compared to only 5 when using wide margins. The width of the margins had in general a greater impact on DVH data for the rectum than for the intestine, apart from at the lowest dose level (Table 1). The fractional rectum volumes receiving the highest doses ðvRect;73% 2 vRect;100% Þ increased by factors in the range 3.6 to 5.0 when using wide rather than narrow margins. The rectum DVH constraints in the intermediate to high dose range were easily fulfilled, also when using the widest margins. A similar impact of the choice of margin was found for the intestine and rectum NTCPs with different fractionation alternatives (Table 2). In general, the intestine NTCPs were higher than the rectum NTCPs for the same margin and fractionation combination. When using the probit model with the Burman et al. data for the intestine, our calculations suggested that using wide as opposed to narrow margins increased the intestine NTCPs with factors in the range of 2.4 to 2.6 for the different fractionation alternatives. As in the DVH analysis, the impact of margin size on NTCPs was stronger for the rectum than for the intestine. Still the impact of margins depended strongly on the seriality parameter of the NTCP model. With wide Table 1 A comparison of intestine and rectum DVH-data using narrow and wide CTV –PTV margins applying standard 3- and 4-field treatment techniques. Average absolute intestine volumes and fractional rectum volumes above various relative dose levels, mostly corresponding to intestine and rectum DVH constraints, are shown. The fractions of patients fulfilling DVH constraints scaled to the 64 Gy/32 fractionation schedule are given in brackets Narrow margins
Wide margins
3-Field
4-Field
3-Field
4-Field
Intestine VInt;48% (cm3) VInt;77% (cm3) VInt;84% (cm3) VInt;100% (cm3)
85.5 (12/15) 61.3 (11/15) 56.6 (2/15) 31.1
102.4 (11/15) 61.5 (11/15) 56.6 (2/15) 27.7
132.7 (10/15) 108.0 (5/15) 102.9 (0/15) 69.2
153.4 (8/15) 108.2 (5/15) 103.5 (0/15) 67.2
Rectum VRect;47% (%) VRect;73% (%) VRect;88% (%) VRect;100% (%)
24.3 6.0 (15/15) 4.7 (15/15) 2.7 (15/15)
81.0 6.9 (15/15) 4.8 (15/15) 2.6 (15/15)
52.3 22.3 (15/15) 17.3 (15/15) 12.3 (14/15)
89.3 24.7 (15/15) 18.2 (15/15) 13.1 (14/15)
68
L.P. Muren et al. / Radiotherapy and Oncology 71 (2004) 65–71
Table 2 Mean and standard deviation (in brackets) of the intestine and rectum NTCPs (%) comparing treatment with combinations of narrow and wide CTV– PTV margins and various fractionation schedules with the 3-field technique 52.5 Gy/20 fxs
55 Gy/20 fxs
64 Gy/32 fxs
57.5 Gy/20 fxs
Narrow margins Intestine Probit model and Burman data
1.1 (1.3)
2.1 (2.4)
2.3 (2.6)
4.0 (4.3)
Rectum Probit model and Burman data ðn ¼ 0:12Þ Probit model and modified Burman data ðn ¼ 0:06Þ Probit model and modified Burman data ðn ¼ 0:20Þ
0.1 (0.2) 0.5 (0.7) 0.01 (0.04)
0.2 (0.4) 1.1 (1.5) 0.04 (0.07)
0.3 (0.4) 1.2 (1.6) 0.04 (0.07)
0.5 (0.9) 2.3 (3.0) 0.09 (0.2)
Wide margins Intestine Probit model and Burman data
2.9 (2.8)
5.3 (2.4)
5.8 (2.6)
9.5 (4.3)
Rectum Probit model and Burman data ðn ¼ 0:12Þ Probit model and modified Burman data ðn ¼ 0:06Þ Probit model and modified Burman data ðn ¼ 0:20Þ
0.8 (1.0) 1.9 (1.8) 0.3 (0.5)
1.6 (2.0) 3.9 (3.5) 0.6 (1.0)
1.8 (2.2) 4.2 (3.7) 0.7 (1.2)
3.2 (3.9) 7.5 (6.4) 1.3 (2.1)
margins the average NTCPs increased by a factor of 6 to 8 when using the standard Burman parameters, compared to factors in the ranges of 3 to 4 and 15 to 30 with modified seriality, n ¼ 0:06 and n ¼ 0:20; respectively. A strong correlation between individual DVH data and NTCP predictions was observed for intestine and rectum volumes above all dose levels assessed (Figs. 1 and 2). The volumes above the highest doses had a slightly stronger impact on the NTCP estimates (i.e., a larger gradient for the straight line fit) than the volumes above the lower dose level. 3.2. The impact of treatment fractionation The average NTCPs for both organs varied in general by factors in the range 3 to 6 across the different fractionation schedules (Table 2). A factor of approximately 9 was observed for one of the modified probit model parameterisations for the rectum ðn ¼ 0:20Þ; but this parameterisation resulted in very low average NTCP estimates. The 64 Gy/32 and the 55 Gy/20 schedules produced very similar NTCPs both for intestine and rectum. For clinics using narrow margins and a fractionation of 52.5 Gy/20, the NTCP modelling suggested that a change in fractionation schedule to 55 Gy/20 or 64 Gy/32 or a change to wide margin would have a similar effect on the intestine NTCP predictions.
4. Discussion This study compared intestine and rectum DVH data and NTCPs in bladder irradiation with a range of margins and fractionation schedule combinations. It was conducted using two widely used beam configurations, i.e. three-field
and four-field CRT arrangements, which led to very similar intestine and rectum irradiation (Table 1). As a consequence, the average NTCPs for the two techniques differed by less than 0.2% (probit model and standard Burman parameters, on all combinations of the 52.5 Gy/20 and the 64 Gy/32 schedules and narrow and wide margins; data not shown). On the other hand, the choice of margins and fractionation schedule was shown to cause large differences in the normal tissue doses/volumes. The choice of treatment margin increased the absolute volumes of intestine receiving a high dose by approximately a factor of two, while the fractional rectum volumes receiving a high dose increased even more, by approximately a factor of four (Table 1). This reflects the close proximity of the rectum to the treatment volume and the relatively small axial extent of the rectum. Still, scoring the fulfilment of a/b-corrected intestine and rectum DVH constraints (64 Gy/32 fractionation used) confirmed that the intestine is the major organ at risk in bladder irradiation, in line with clinical experience [24] and previous NTCP modelling [18]. The results of the NTCP analysis comparing the impact of variation in treatment margins (i.e. volume effects) and treatment fractionation (i.e. dose effects) depended heavily on the radiation sensitivity parameter of the NTCP model. Probably due to difficulties caused by the shape and mobility of the intestine, there has been few NTCP modelling studies for this organ. With the probit model we therefore only applied the standard Burman et al. data. Using these parameters, however, rather similar findings in relation to the impact of the size of margins were found in the NTCP modelling and the DVH analysis. The importance of the seriality parameter was clearly emphasised in the rectum NTCP analysis, as the calculations were repeated with three different n parameters, altering the volume effect/seriality. When using parameters expressing low
L.P. Muren et al. / Radiotherapy and Oncology 71 (2004) 65–71
69
Fig. 1. The correlation between four sets of intestine DVH-data from each of the 15 individual patients and their corresponding intestine NTCP estimate, using the probit model with the standard Burman parameters to evaluate the treatment data for the 64 Gy/32 fractionation schedule, wide margins and the 3-field technique. The linear regression curves were derived from the individual DVH-data for each of the dose levels. A strong correlation between DVH data and NTCP estimates was observed for the volumes above all dose levels. The volumes above the highest doses had a slightly stronger impact on the NTCP estimates (i.e., a larger gradient for the straight line fit) than the volumes above the lower dose level. The literature constraints at the intermediate dose levels are in agreement with the Burman parameters for the probit model, as they fall at an NTCP level of approximately 5%.
volume effect/high seriality ðn ¼ 0:06Þ; the variation in NTCPs between different margins was relatively low and similar to the variation between different fractionation schedules. However, for the standard Burman parameters and the parameterisation reflecting increased volume effect/lower seriality ðn ¼ 0:20Þ; the choice of margin caused considerably larger variation in the NTCPs than the choice of fractionation. Good agreement was observed also at the individual level between DVH and NTCP data (Figs. 1 and 2). Based on these relations, it is possible to estimate within 1% the predicted NTCP using one DVH parameter only. In addition, these relations can be used to derive the expected increase in NTCP after a change in the volume receiving at least a certain dose level. Intestine and rectum literature constraints were also included in Figs. 1 and 2. For the intestine, two of the constraints (at the 48 and 77% dose levels) were in agreement with the applied NTCP parameters and model since these constraints were found at an NTCP level of approximately 5%. The intestine constraint at the higher
dose level was very strict, both in the sense that it corresponded to a negligible NTCP level and in the sense that it was fulfilled in only a few of the cases analysed (Table 1). This might indicate that there is a need for updated, prospectively recorded intestine follow-up data, in particular for higher dose levels. For the rectum, all literature constraints were in agreement with the standard Burman parameters for the probit NTCP model, as they were located at an NTCP level of approximately 5%. When using the two modified parameterisations, the agreement between the NTCP model and the literature constraints was less good (data not shown). The NTCP data presented here did not allow any validation of the applied NTCP model. Nevertheless the average NTCPs appear to be in line with observed late adverse effect rates for both previous and current margin and fractionation combinations at WGH and HUH [8,26, 27]. For instance, the 57.5 Gy/20 schedule—though with standardised (typically 10 £ 10 cm2) field sizes—was abandoned due to a high intestinal adverse effect rate [8, 21]. Obviously, we advocate cautious interpretation of
70
L.P. Muren et al. / Radiotherapy and Oncology 71 (2004) 65–71
Fig. 2. The correlation between four sets of rectum DVH-data from each of the 15 individual patients and their corresponding rectum NTCP estimate, using the probit model with the standard Burman parameters (same treatment technique, margins and fractionation as in Fig. 1). The linear regression curves were derived from the individual DVH-data for each of the dose levels. A strong association between DVH data and NTCP estimates was observed also for the rectum, at least for the volumes above the three highest dose levels. All literature constraints were in agreement with the standard Burman parameters for the probit model, as they fall at an NTCP level of approximately 5%.
NTCP modelling. It still seems that clinics that irradiate bladder cancer using narrow margins and the 52.5 Gy/20 schedule may consider an increase in the margins to better account for geometrical uncertainties (bladder motion and patient positioning) or a moderate dose escalation. The current analysis suggests that the 55 Gy/20 and the 64 Gy/ 32 schedules are equivalent in terms of intestine and rectum irradiation. Calculations on a subset of the patients also confirmed that the 57.5 Gy/20 and a 68 Gy/34 schedule resulted in almost identical intestine and rectum NTCP estimates (data not shown). With the clinical experience of the 57.5 Gy/20 fractionation in mind, this also shows that escalation of the dose to the whole bladder above these levels should not be investigated without implementing highly conformal treatment techniques, such as intensity modulated RT (IMRT), that better spare the involved normal tissues. Treatment approaches such as IMRT and/or imageguided RT are currently being introduced to improved normal tissue sparing and if possible, to allow dose escalation. These techniques are likely to become of considerable benefit also in the management of bladder
cancer [20]. In addition to showing that these techniques probably are required for further dose escalation without increased morbidity in bladder irradiation, this study showed that large benefits can be realised by optimisation of more fundamental treatment parameters, such as the width of treatment margins and the choice of fractionation. This potential should be exploited before pursuing the more technologically sophisticated treatment approaches. In conclusion, this modelling study of bladder irradiation suggested that the choice of margins was as important as the choice of fractionation in terms of intestine and rectum DVH parameters and NTCP predictions. The 55 Gy/20 and 64 Gy/32 fractionation schedules appeared to be comparable in terms of intestine and rectum NTCP predictions.
Acknowledgements Generous grants from the Norwegian Cancer Society, The Norwegian foundation for Health and Rehabilitation, the University of Bergen and Haukeland University Hospital supported this study. The authors want to acknowledge the
L.P. Muren et al. / Radiotherapy and Oncology 71 (2004) 65–71
support from Olav Dahl, August Bakke, Jarle Rørvik and Ole Johan Halvorsen at Haukeland University Hospital and David Thwaites and Grahame Howard at Western General Hospital for the conduction of this study.
References [1] Baglan KL, Frazier RC, Yan D, et al. The dose–volume relationship of acute small bowel toxicity from concurrent 5-FU-based chemotherapy and radiation therapy for rectal cancer. Int J Radiat Oncol Biol Phys 2002;52:176–83. [2] Boersma LJ, van den Brink M, Bruce AM, et al. Estimation of the incidence of late bladder and rectum complications after high-dose (70–78 GY) conformal radiotherapy for prostate cancer, using dose– volume histograms. Int J Radiat Oncol Biol Phys 1998;41:83– 92. [3] Borgaonkar S, Jain A, Bollina P, et al. Radical radiotherapy and salvage cystectomy as the primary management of transitional cell carcinoma of the bladder. Results following the introduction of a CT planning technique. Clin Oncol (R Coll Radiol) 2002;14:141– 7. [4] Burman C, Kutcher G, Emami B, Goitein M. Fitting of normal tissue tolerance data to an analytic function. Int J Radiat Oncol Biol Phys 1991;21:123– 35. [5] Chahal R, Sundaram SK, Iddenden R, et al. A study of the morbidity, mortality and long-term survival following radical cystectomy and radical radiotherapy in the treatment of invasive bladder cancer in Yorkshire. Eur Urol 2003;43:246 –57. [6] Dale E, Hellebust TP, Skjønsberg A, Høgberg T, Olsen DR. Modeling normal tissue complication probability from repetitive CT scans during fractionated HDR brachytherapy and external beam radiotherapy of the uterine cervix. Int J Radiat Oncol Biol Phys 2000;47: 963–71. [7] Dale E, Olsen DR, Fossa˚ SD. Normal tissue complication probabilities correlated with late effects in the rectum after prostate conformal radiotherapy. Int J Radiat Oncol Biol Phys 1999;43:385–91. [8] Duncan W, Quilty PM. The results of a series of 963 patients with transitional cell carcinoma of the urinary bladder primarily treated by radical megavoltage X-ray therapy. Radiother Oncol 1986;7: 299–310. [9] Emami B, Lyman J, Brown A, et al. Tolerance of normal tissue to therapeutic irradiation. Int J Radiat Oncol Biol Phys 1991;21:109–22. [10] Gallagher MJ, Brereton HD, Rostock RA, et al. A prospective study of treatment techniques to minimize the volume of pelvic small bowel with reduction of acute and late effects associated with pelvic irradiation. Int J Radiat Oncol Biol Phys 1986;12:1565–73. [11] Jackson A, Skwarchuk MW, Zelefsky MJ, 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. [12] Kutcher GJ, Burman C. Calculation of complication probability factors for non-uniform normal tissue irradiation: the effective volume method. Int J Radiat Oncol Biol Phys 1989;16:1623–30.
71
˚ gren A, Brahme A. Tumor and normal tissue responses [13] Ka¨llman P, A to fractionated non-uniform dose delivery. Int J Radiat Biol 1992;62: 249–62. [14] Lebesque JV, Bruce AM, Kroes AP, et al. Variation in volumes, dosevolume histograms, and estimated normal tissue complication probabilities of rectum and bladder during conformal radiotherapy of T3 prostate cancer. Int J Radiat Oncol Biol Phys 1995;33:1109– 19. [15] Lyman JT. Complication probability as assessed from dose volume histograms. Radiat Res 1985;104:S13–S19. [16] Marks LB, Carroll PR, Dugan TC, Anscher MS. The response of the urinary bladder, urethra, and ureter to radiation and chemotherapy. Int J Radiat Oncol Biol Phys 1995;31:1257 –80. [17] Meijer GJ, Rasch C, Remeijer P, Lebesque JV. 3D analysis of delineation errors, setup errors and organ motion during radiation therapy of bladder cancer. Int J Radiat Oncol Biol Phys 2003;55: 1277– 87. [18] Muren LP, Jebsen N, Gustafsson A, Dahl O. Can dose response models predict reliable normal tissue complication probabilities in radical radiotherapy of urinary bladder cancer? The impact of alternative radiation tolerance models and parameters. Int J Radiat Oncol Biol Phys 2001;50:627– 37. [19] Muren LP, Smaaland R, Dahl O. Organ motion, set-up variation and treatment margins in radical radiotherapy of urinary bladder cancer. Radiother Oncol 2003;69(3):291 –304. [20] Muren LP, Smaaland R, Dahl O. Recent advances and future challenges in conformal radiotherapy of muscle-invading urinary bladder cancer. Radiother Oncol. Conditionally accepted. [21] Quilty PM, Duncan W, Kerr GR. Results of a randomised study to evaluate influence of dose on morbidity in radiotherapy for bladder cancer. Clin Radiol 1985;36:615–8. [22] Redpath AT. A beam model for three-dimensional radiotherapy planning. Br J Radiol 1995;68:1356–63. [23] Redpath AT. Planning of beam intensity modulation using an advanced 3D dose calculation algorithm and a simulated annealing method. Radiother Oncol 1998;49:295–304. [24] Sengelo¨v L, von der Maase H. Radiotherapy in bladder cancer. Radiother Oncol 1999;52:1–14. [25] Skwarchuk MW, Jackson A, Zelefsky MJ, et al. Late rectal toxicity after conformal radiotherapy of prostate cancer (I): multivariate analysis and dose-response. Int J Radiat Oncol Biol Phys 2000;47: 103–13. [26] Sma˚land R, Akslen LA, Tønder B, et al. Radical radiation treatment of invasive and locally advanced bladder carcinoma in elderly patients. Br J Urol 1991;67:61–9. [27] Tait DM, Nahum AE, Meyer LC, et al. Acute toxicity in pelvic radiotherapy; a randomised trial of conformal versus conventional treatment. Radiother Oncol 1997;42:121–36. [28] Ten Haken R. Introduction. Semin Radiat Oncol 2001;11:181– 3. [29] Zietman AL, Shipley WU. Organ-sparing treatment for bladder cancer: time to beat the drum. Int J Radiat Oncol Biol Phys 1994;30: 741–2.