Download determination of optimum planning target volume margins for

236
Exp Oncol 2004
26, 3, 236-239
Experimental Oncology 26, 236-239, 2004 (September)
DETERMINATION OF OPTIMUM PLANNING TARGET VOLUME
MARGINS FOR VARIOUS TUMOR SITES USING ELECTRONIC
PORTAL IMAGING
M. Nalca Andrieu1,*, B. Dirican2, A.Y. Ozturk1, A. Hicsonmez1, C. Kurtman1
Department of Radiation Oncology, Ankara University Medical School, Ankara, Turkey
2
Department of Radiation Oncology, Gulhane Military Medical Academy, Ankara, Turkey
1
Objective: The aim of this study is to evaluate the efficacy of electronic portal imaging (EPI) to measure the setup errors for four different sites of irradiation caused by patient positioning. Methods: A total number of 95 portal
images of 11 patients (3 pelvic, 1 total cranium, 3 mantle and 4 tangential fields for breast) were collected during
the course of study. The first portal images after a correction of set-up errors according to the simulation films were
accepted as the reference images for the subsequent sessions. By matching each portal image with the reference
image, the deviations in lateral (x) and superior-inferior (y) axis for all and additionally in antero-posterior (z) axis
for pelvis, and standard deviations were calculated. Results: The set-up errors caused by patient’s positioning are
completely abolished in 15 mm planning target volume (PTV) margins for all studied cases. Conclusion: Standard
PTV margins usually completely cover the set-up errors caused by patient’s positioning.
Key Words: radiotherapy, portal imaging, PTV.
The goal of treatment planning in radiotherapy is to
spare the surrounding normal tissues while giving the
maximum dose to the tumor tissue. However, any error between the planned and delivered treatment positions is likely to result in degradation of the therapeutic ratio, with the possibility of underdosing the target or
overdosing the normal tissues. The target, as defined
here, is the planning target volume (PTV), which includes the clinical target volume (CTV) and a margin to
account for treatment uncertainties [6, 7]. The CTV must
encompass the gross tumor volume (GTV) with subclinical disease and possibly involved lymph nodes. The
margins of PTV must guarantee adequate coverage of
the CTV during the treatment to avoid target misses
and local recurrence, whilst margins that are too large
may lead to other complications and morbidity. Addition of margins around the CTV is one of the methods
to overcome the difficulties in isocenter targeting. In
order to determine these margins, it is useful to think of
the two types of uncertainties, internal and set-up margins (SMs). The internal margin (IM) must be added to
the CTV to compensate for expected physiologic movements and variations in size, shape and position of the
CTV during therapy in relation to an internal reference
point and its corresponding coordinate systems. These
internal variations do not depend on external uncertainties in beam geometry, but could depend on patient day-to-day set-up. On the other hand, the SM
accounts specifically for uncertainties in patient positioning and alignment of the therapeutic beams during
Received: April 2, 2004.
*Correspondence:
Fax: +90 312 4465912; +90 312 4371945
E-mail : [email protected]
Abbreviations used: CTV — clinical target volume; EPI — electronic portal imaging; IM — internal margin; PTV — planning target volume; SM — set-up margin.
treatment planning and through all treatment sessions,
and is referenced in the external coordinate system [7].
There may be several errors during the delivery of the
treatment due to the organ motions, inadequate beam
placement and inaccurate patient positioning. It is not
possible to eliminate all these errors, but an understanding of the source and magnitude of these errors
is essential to reduce these errors and to determine
the adequacy of the applied margins [5].
Electronic portal imaging devices (EPID) are increasingly used in clinical practice to improve set-up
accuracy. Clinical studies comparing portal films with
electronic portal images (EPI) suggest that EPI is a
suitable replacement for films [11, 12]. In comparison
with portal films, acquisition is faster and the evaluation is greatly facilitated by using computerized analysis methods. The verification of patient position is commonly performed on the first day of treatment and thereafter on a weekly basis. The portal images are
compared with the planned treatment position represented by a set of reference images, namely simulator
films, initial portal images, or digitally reconstructed radiographs (DRRs) [4]. This makes it possible to correct
set-up errors before completing a treatment (on-line)
or before the next treatment (off-line).
The aim of this study is to apply portal imaging to
measure the set-up errors for different sites of irradiation caused by organ motion and inaccurate patient
positioning.
Patient positioning. A total number of 95 portal
images of 11 patients with different cancer sites were
used in this study. The characteristics of tumors and
treatment details are shown in Table 1. The treatments
involved 3 pelvic field patients with box technique, 1
total cranium with two lateral fields, 3 mantle irradiation
with two opposing antero-posterior fields and 4 breast
Experimental Oncology 26, 236-239, 2004 (September)
237
7DEOH&KDUDFWHULVWLFVRIWXPRUVDQGWUHDWPHQW
7XPRUORFDOL]DWLRQ
+LVWRORJLFDOW\SH
6WDJH
2SHUDWLRQ
5DGLRWKHUDS\
5HFWXP
$GHQRFDUFLQRPD
710
/$5
$GMXYDQW
5HFWXP
$GHQRFDUFLQRPD
710
$35
$GMXYDQW
(QGRPHWULXP
$GHQRFDUFLQRPD
,&
7$+%62
$GMXYDQW
%UDLQ
(SLGHUPRLG&DPHW
,9
£
3DOLDWLYH
&HUYLFDOO\PSKQRGH
/3+'
,$
£
3ULPDU\FXUDWLYH
&HUYLFDOO\PSKQRGH
16+'
,$
£
3ULPDU\FXUDWLYH
&HUYLFDOPHGLDVWLQDOO\PSKQRGH
16+'
,,$
£
3ULPDU\FXUDWLYH
/HIWEUHDVWRXWHUTXDGUDQW
,QYDVLYHGXFWDO&D
710
050
$GMXYDQW
5LJKWEUHDVWRXWHUTXDGUDQW
,QYDVLYHGXFWDO&D
710
050
$GMXYDQW
5LJKWEUHDVWRXWHUTXDGUDQW
,QYDVLYHGXFWDO&D
710
/XPSHFWRP\
$GMXYDQW
5LJKWEUHDVWRXWHUTXDGUDQW
,QYDVLYHGXFWDO&D
710
/XPSHFWRP\
$GMXYDQW
1RWHV$35£DEGRPLQRSHULQHDOUHVHFWLRQ%62£ELODWHUDOVDOSLQJRRRSKRUHFWRP\/$5£ORZDQWHULRUUHVHFWLRQ/3+'£O\PSKRF\WHSUHGRPLQDQW
+RGJNLQGLVHDVH050£PRGLILHGUDGLFDOPDVWHFWRP\16+'£PRGXODUVFOHURVLQJ+RGJNLQGLVHDVH7$+£WRWDODEGRPLQDOK\VWHUHFWRP\
cancer patients (2 with mastectomy and 2 with conservative surgery) with tangential fields. The treatment
fields were applied to all of the patients according to
standard protocols that were not directly related with
the primary tumor size. Eight patients were simulated
and treated in the supine position and 3 patients with
pelvic fields were in the prone position. Patient positioning was achieved by laser adjustment to skin marks
and immobilization devices were used for each of them.
Alpha-cradles were used for pelvis and mantle fields,
thermoplastic mask was used for cranium and breast
board for tangential irradiations, additionally we used
breast ring for one of the patients who had conservative breast surgery for pendulous breast. Simulation
films were taken after the treatment portals had been
simulated and the accuracy of the field placement plots
had been judged by the radiation oncologist.
Portal imaging and matching. The EPIs were taken
under 6 MV photon exposure of Varian (Varian Associates Inc., Oncology Systems, Palo Alto, CA, USA) Clinac 2300 C/D linear accelerator. The simulator images
were digitized using a charge coupled device camera
and a light table and imported to image registration software for analysis, using the built-in important feature.
After the simulator image was digitized by means of the
Portal Vision’s built in frame grabber from a camera, the
reference images were tagged so that the matching software recognized them as reference images. The procedure was carried out using Varian Portal Vision software (v.3.4). The first image that was taken for the reference EPI was scaled according to the information given
before as known and measured marker distance. The
first portal images which were taken after a correction of
set-up errors by comparing with the simulation films were
accepted as the reference images for the subsequent
sessions. These reference images were used for image
matching and on-line set-up verification. Patient setup errors were determined by comparing the patient’s
portal films, representing the actual treatment position,
with the corresponding simulator film that represents the
planned treatment position. The intended field edge was
determined by drawing a polygon around the field edge
that was given on the reference image by the wires in
the reference image. Then anatomical landmarks and/or
bone or other visible anatomical contours were defined.
The program estimated their position using the intended
field edge (defined in the reference image) and the detected field edge (extracted from the portal image).
A known distance like a distance marker or field size in
the reference image was measured during the drawing
of the reference structures. These values were used in
calculating the scale factor of the reference image and
this value was used later for measuring set-up errors as
isocentric distances in the portal image. Mouse driven
measurements of distances were useful for finding out
placement errors. Bony landmarks selected from the
reference image, were superimposed and aligned onto
the portal image. The detected field edges of all matched
portal images over the reference image and their respective field edge and anatomical landmarks or contours were displayed. The anatomic landmark displacements from the reference image were determined in the
lateral (x) and superior-inferior (y) directions for all cranial, breast, mantle and pelvic treatment fields and additional the antero-posterior direction (z) was used for
pelvic box treatment. Isocenter adjustments were performed accordingly for displacement errors and treatment was continued. Following the set-up verification,
treatment of the remaining fields was continued using 6
MV photons except for the pelvic fields that were additionally treated with a boosted dose (using 20 MV photons) to compensate for the 6 MV treatment administered during the first field of the fraction.
Analysis of set-up errors. The set-up variability
of all patients was characterized by the standard deviation (SD) of the displacement of the reference point of
the portal image from the reference point of the corresponding simulator film for the specified coordinate as
described by equation 1:
¦¦ '
3
' (eq. 1)
Q where s is the SD, P the individual patient, D the
displacement of the reference point of portal image from
the corresponding simulator film, and ' the mean displacement of all images in the observed direction [9].
Additionally, standard deviations of the displacement
of the reference point of the portal image from the reference point of the corresponding reference EPI after
corrections were also calculated (SD’).
The volume data was evaluated by only x and y
axis whereas in the cranial and mantle opposing field
and breast tangential irradiations, the treatment uncertainties regarding the third dimension displacement can
be assumed as neglegible. For these cases, only the
V
3
L
L
238
Experimental Oncology 26, 236-239, 2004 (September)
displacements in two directions would affect the coverage volume. Neverthless, pelvic fields have been irradiated with four field box technique that z axis was
added for pelvic sites. The CTV contours in CT sections or orthogonal simulation films were transferred to
the treatment planning system (CadPlan) using a digitizer. The isodoses were obtained after the optimum
treatment planning for each tumor site. The minimum
coverage was defined as the smallest amount of coverage that at least 95% of the volume of the immobile
target had received. The relationship between margin
increase and coverage was measured by applying increasing theoretical margins of 5, 10 and 15 mm to our
distributions of coverage and determining the minimum
coverage value inside the target. The minimum coverage of at least 95% volume coverage was found out for
each treatment by using the 95% isodoses in the treatment planning system.
Standard deviations of the displacements of the
reference point of the portal images from the reference
point of the corresponding simulator films before corrections (SD) and displacement from the reference point
of the reference images after corrections (SD’) are given
with the mean displacement values ( ' ) in Table 2 for
pelvis, brain, mantle and breast fields, respectively. The
percentage of the treatments with minimum 95% volume coverage taken by 5 mm, 10 mm, and 15 mm
margins of PTV are also given in Table 3. Random errors that reflect the organ motions and daily set-up errors in the superior-inferior (y) direction were frequently
larger in magnitude than those in the lateral (x) direction for brain, mantle and breast fields whilst y direction
values was less than that of x direction for pelvic fields.
The differences between 95% volume coverage rates
for 5 mm, 10 mm, and 15 mm margins showed an inhomogenity of the dose inside PTV. This inhomogenity
was probably related with the irradiation techniques
used with beam modifiers. Especially, for tangential irradiations, the effect of organ motion and discrepancies of patient positioning may limit the better dose
distribution expected for wedged fields also in connection with the body contour.
Several studies have been previously done to evaluate the occurrence of isocenter set-up errors using a
variety of different imaging techniques. As there are
too many differences regarding the data of tumor sites
and treatment techniques studied in the literature, it is
difficult to compare our results with the others. In our
7DEOH&KDUDFWHULVWLFVRIWKHVHWXSYDULDELOLW\RIWKHSDWLHQWV
/RFDOL]DWLRQRI
WKHWXPRU
3HOYLV
%UDLQ
0DQWOHILHOG
%UHDVW
[PP
–
6' 6'
'
\PP
]PP
–
–
6' 6'
' 6' 6'
'
7DEOH9ROXPHFRYHUDJHUDWHVIRUDQGPPPDUJLQV
/RFDOL]DWLRQ
3HOYLV
%UDLQ
0DQWOH
%UHDVW
0DUJLQVPP
study, considering the SD according to set-up errors,
5.0 mm of PTV margins is sufficient only for cranial
treatments to ensure 95% coverage. Depending on the
clinical site of irradiation, the PTV could be very similar
to the CTV. Our findings also supported this as in the
IM approached to a very low value for brain tumors
due to a good immobilization technique and the lack of
physiologic movements. By contrast, the PTV could be
much larger for breast, lung and pelvic tumors with the
contribution of organ motions and difficulties of patient
positioning. In the present study, it was observed that
for pelvic, mantle and tangential breast fields 15 mm of
margins might be safe enough unless additional setup corrections and isocenter adjustments are applied
routinely.
There have been several studies defining field margins for the construction of PTV from a CTV-using data
on isocenter set-up error for prostate treatments [1, 2,
8, 10]. Alasti and his colleagues [1] used portal imaging to measure daily on-line set-up error and off-line
prostatic motion in patients treated with conformal radiotherapy to determine an optimum PTV margin. They
suggested PTV margins of 10.0 mm in the antero-posterior (AP) direction and 5.9 mm in the superior-inferior (SI) direction to ensure 95% coverage. These findings were loosely matching those of Antolak et al [2]
who calculated 95% coverage margins of 11.0 mm AP
and 7.0 mm SI using similar techniques. Tinger et al
[10] have used both set-up error and organ motion,
and calculated margins of 8.0 mm AP and 8.8 mm SI
by using two standard deviations to account for 95% of
the uncertainties. Melian et al [8] used one-tailed, onedimensional cumulative probability to arrive at a 90%
coverage margin of 7.0 mm. However, we have to keep
in mind that all of these studies were taking into account organ motion and set-up errors in relatively
smaller pelvic fields, and mostly using 3-D conformal
techniques that was not the case in our study.
In several other studies, the accuracy of daily setups has been assessed using EPIs in patients with prostate cancer. In the retrospective analysis of Hanley et al
[5] which includes a large number of portal images, the
set-up verification protocol appeared to minimize systematic set-up errors to a level that approaches the sensitivity of the image registration technique. The random
day-to-day fluctuations, represented by the average
values of the standard deviations, were minor in comparison to the currently used margins. Balter et al [3]
concluded that the use of on-line imaging and image
registration to guide adjustment of patient set-up may
lead to a reduction in the volume of normal tissues irradiated and possibly improve the probability of complication-free survival in future treatments.
In conclusion, we can say that the verification of
treatment portals by using EPI is feasible and useful.
Application of routine set-up verification protocol may
lead to a reduction in PTV margins. However, PTV
margins in the first hand strongly depend on clinical
indications, and the magnitude of this potential reduction should be investigated in large number of patients.
Experimental Oncology 26, 236-239, 2004 (September)
REFERENCES
1. Alasti H, Petric MP, Catton CN, Warde PR. Portal imaging for evaluation of daily on-line setup errors and
off-line organ motion during conformal irradiation of carcinoma of the prostate. Int J Radiat Oncol Biol Phys 2001;
49: 869–84.
2. Antolak JA, Rosen II, Childress CH, Zagars GK,
Pollack A. Prostate target volume variations during a
course of radiotherapy. Int J Radiat Oncol Biol Phys 1998;
42: 661–72.
3. Balter JM, Chen GTY, Pelizzari CA, Krishnasamy S,
Rubin S, Vijayakumar S. Online repositioning during treatment of the prostate: A study of potential limits and gains.
Int J Radiat Oncol Biol Phys1993; 27: 137–43.
4. Bel A, Vos PH, Rodrigus PTR, Creutzberg CL,
Visser AG, Stroom JC. High-precision prostate cancer
irradiation by clinical application of an off-line patient
set-up verification procedure, using portal imaging. Int J
Radiat Oncol Biol Phys 1996; 35: 321–32.
5. Hanley J, Lumley MA, Mageras GS, Sun J, Zelefsky MJ, Leibel SA. Measurement of patient positioning
errors in three-dimensional conformal radiotherapy of the
prostate. Int J Radiat Oncol Biol Phys 1997; 37: 435–44.
6. ICRU Report No.50, Prescribing, recording and reporting photon beam therapy, International Commission on
Radiation Units and Measurements 1993; Bethesda, MD.
239
7. ICRU Report No.62, Prescribing, recording and reporting photon beam therapy (supplement to ICRU Report 50), International Commission on Radiation Units
and Measurements 1999, Bethesda, MD.
8. Melian E, Mageras GS, Fuks Z, Leibel SA, Niehaus A, Lorant H. Variation in prostate position quantitation and implications for three-dimensional conformal
treatment planning. Int J Radiat Oncol Biol Phys 1997;
38: 73–81.
9. Rudat V, Schraube P, Oetzel D, Zierhut D, Flentje M,
Wannenmacher M. Combined error of patient positioning
variability and prostate motion uncertainty in 3D conformal
radiotherapy of localized prostate cancer. Int J Radiat Oncol
Biol 1996; 35: 1027–34.
10. Tinger A, Michalski JM, Cheng A, Low DA, Zhu R,
Bosch WR. A critical evaluation of the planning target
volume for 3-D conformal radiotherapy of prostate cancer.
Int J Radiat Oncol Biol Phys 1998; 42: 213–21.
11. Thompson V, Bidmead M, Mubata C. Pictorial
review: comparison of portal imaging and megavoltage verification films for conformal pelvic irradiation. Br J Radiol
1996; 69: 1191–5.
12. Yin FF, Rubin P, Schell MC, Wynn R, Raubertas RF, Uschold G. An observer study for direct comparison of clinical efficacy of electronic to portal images. Int J
Radiat Oncol Biol Phys 1996; 35: 985–91.
ÎÏÐÅÄÅËÅÍÈÅ ÎÏÒÈÌÀËÜÍÎÃÎ ÏËÀÍÈÐÓÅÌÎÃÎ ÎÁÚÅÌÀ
ÌÈØÅÍÈ ÎÁËÓ×ÅÍÈß ÄËß ÎÏÓÕÎËÅÉ ÐÀÇËÈ×ÍÎÉ
ËÎÊÀËÈÇÀÖÈÈ Ñ ÈÑÏÎËÜÇÎÂÀÍÈÅÌ ÝËÅÊÒÐÎÍÍÎÉ
ÏÎÐÒÀËÜÍÎÉ ÂÈÇÓÀËÈÇÀÖÈÈ
Öåëü: îöåíèòü ïîãðåøíîñòü, ñâÿçàííóþ ñ ïîëîæåíèåì ïàöèåíòà ïðè îáëó÷åíèè îïóõîëè, ñ ïîìîùüþ ýëåêòðîííîé
ïîðòàëüíîé âèçóàëèçàöèè (EPI). Ìåòîäû: ïîëó÷åíî 95 ïîðòàëüíûõ èçîáðàæåíèé ó 11 ïàöèåíòîâ ñ ëîêàëèçàöèåé
îïóõîëè â îáëàñòè òàçà (3), ÷åðåïà (1), êîðû ãîëîâíîãî ìîçãà (3) è ìîëî÷íîé æåëåçû (4). Ïåðâûå ïîðòàëüíûå
èçîáðàæåíèÿ ïîñëå êîððåêòèðîâêè ïîãðåøíîñòåé íà ñèìóëÿòîðå ïðèíèìàëè çà ñòàíäàðò äëÿ ïîñëåäóþùåãî àíàëèçà.
Ïóòåì ñðàâíåíèÿ êàæäîãî ïîðòàëüíîãî èçîáðàæåíèÿ ñî ñòàíäàðòíûì äëÿ êàæäîãî ñåàíñà îáëó÷åíèÿ áûëè âû÷èñëåíû
ðàçáðîñû äëÿ ëàòåðàëüíûõ (x) è âíåøíèõ-âíóòðåííèõ (y) îñåé, à êðîìå òîãî äëÿ òàçîâîé îáëàñòè — äîïîëíèòåëüíî
ïî ïåðåäíèì-çàäíèì (z) îñÿì, ïîñëå ÷åãî áûëè ïðîñ÷èòàíû ñòàíäàðòíûå îòêëîíåíèÿ. Ðåçóëüòàòû: äëÿ
èññëåäîâàííûõ ëîêàëèçàöèé 15 ìì ãðàíèöû äëÿ PTV íèâåëèðóþò ïîãðåøíîñòè, ñâÿçàííûå ñ óêëàäêîé ïàöèåíòà.
Âûâîäû: ñòàíäàðòíûå ãðàíèöû PTV îáû÷íî ïîëíîñòüþ ïîêðûâàþò ïîãðåøíîñòè, ñâÿçàííûå ñ óêëàäêîé ïàöèåíòà.
Êëþ÷åâûå ñëîâà: ðàäèîòåðàïèÿ, ïîðòàëüíûé èìèäæèíã, ïëàíèðóåìûé îáúåì îáëó÷åíèÿ ìèøåíè.
Copyright © Experimental Oncology, 2004