Technology in Cancer Research and Treatment

ISSN 1533-0346

Analysis of Dose Distribution and Risk of Danita Kannarunimit, M.D.1

Pneumonitis in Stereotactic Body Radiation Therapy Martina Descovich, Ph.D.2
for Centrally Located Lung Tumors: A Comparison Aaron Garcia, M.S.2
of Robotic Radiosurgery, Helical Tomotherapy and Josephine Chen, Ph.D.2
Volumetric Modulated Arc Therapy Vivian Weinberg, Ph.D.2
Christopher Mcguinness,
Ph.D.2
DOI: 10.7785/tcrt.2012.500394
Dilini Pinnaduwage, Ph.D.2
Stereotactic body radiation therapy (SBRT) to central lung tumors is associated with normal ­tissue John Murnane, Ph.D.2
toxicity. Highly conformal technologies may reduce the risk of complications. This study compares
Alexander R. Gottschalk,
physical dose characteristics and anticipated risks of radiation pneumonitis (RP) among three
SBRT modalities: robotic radiosurgery (RR), helical tomotherapy (HT) and volumetric modulated M.D., Ph.D.2
arc therapy (VMAT). Nine patients with central lung tumors 5 cm were compared. RR, HT and Sue S. Yom, M.D., Ph.D.2*
VMAT plans were developed per RTOG 0831. Dosimetric comparisons included target cover-
age, conformity index, heterogeneity index, gradient index, maximal dose at 2 cm from target
(D2 cm), and dose-volume parameters for organs at risk (OARs). Efficiency endpoints included Department of Radiation Oncology,
1

total beam-on time and monitor units. RP risk was derived from Lyman-­­Kutcher-Burman ­modeling King Chulalongkorn University
on in-house software. The average GTV and PTV were 11.6 6 7.86 cm3 and 36.8 6 18.1 cm3.
Hospital, Bangkok
All techniques resulted in similar target coverage (p 5 0.64) and dose conformity (p 5 0.88).
While RR had sharper fall-off gradient (p 5 0.002) and lower D2 cm (p 5 0.02), HT and VMAT Department of Radiation Oncology,
2

produced greater homogeneity (p  0.001) and delivery efficiency (p 5 0.001). RP risk pre- University of California, San Francisco
dicted from whole or contralateral lung volumes was less than 10%, but was 2-3 times higher
using ipsilateral volumes. Using whole (p 5 0.04, p 5 0.02) or ipsilateral (p 5 0.004, p 5 0.0008)
volumes, RR and VMAT had a lower risk of RP than HT. Using contralateral volumes, RR had
the lowest RP risk (p 5 0.0002, p 5 0.0003 versus HT, VMAT). RR, HT and VMAT were able
to provide clinically acceptable plans following the guidelines provided by RTOG 0813. All tech-
niques provided similar coverage and conformity. RR seemed to produce a lower RP risk for a
scenario of small PTV-OAR overlap and small PTV. VMAT and HT produced greater homoge-
neity, potentially desirable for a large PTV-OAR overlap. VMAT probably yields the lowest RP
risk for a large PTV. Understanding subtle differences among these technologies may assist in
situations where multiple choices of modality are available.

Key words: Central lung; SBRT; Robotic radiosurgery; Helical tomotherapy; VMAT;
Pneumonitis.

Abbreviations: AAA: Analytical Anisotropic Algorithm; BED: Biological Equivalent Dose; CN:
Conformation Number; CI: Conformity Index; DVH: Differential Dose Volume Histogram; DRR:
Digitally Reconstructed Radiographs; FFF: Flattening Filter-Free; gEUD: Generalized Equiva-
lent ­Uniform Dose; GI: Gradient Index; GTV: Gross Tumor Volume; HT: Helical Tomotherapy;
HI: ­Heterogeneity Index; IMRT: Intensity-modulated Radiation Therapy Treatments; ITV: Internal
­Target Volume; CBCT: Cone Beam Computed Tomography; LQ: Linear-Quadratic; MIP: Maximum
­Intensity Projection; MLD: Mean Lung Dose; MVCT: Megavoltage Computed Tomography; MU:
Monitor Unit; MLC: Multileaf Collimator; NSCLC: Non-small Cell Lung Cancer; NTCP: Normal *Corresponding author:
Tissue Complication Probability; OAR: Organ At Risk; PTV: Planning Target volume; RP: Radia- Sue S. Yom, M.D., Ph.D.
tion Pneumonitis; RTOG: Radiation Therapy Oncology Group; RR: Robotic Radiosurgery; SBRT: Phone: 415-353-7175
Stereotactic Body Radiation Therapy; TC: Target Coverage; TCP: Tumor Control Probability; USC: Fax: 415-353-9883
Universal Survival Curve; VMAT: Volumetric Modulated Arc Therapy. E-mail: yoms@radonc.ucsf.edu

49
 Kannarunimit et al.

Introduction Volumetric-modulated arc therapy (VMAT) is a recently

developed radiotherapy optimization technique which
Stereotactic body radiotherapy (SBRT) has become a s­ tandard enables delivery of IMRT treatments by combining continu-
treatment option for medically inoperable early stage non- ously rotating gantry motion with simultaneous variation of
small cell lung cancer (NSCLC) patients, with excellent clin- dose rate, gantry speed and segment shape (26). VMAT treat-
ical outcomes. Previous studies have reported 5-year local ments can be delivered accurately and efficiently in a sub-
control rates of 90% and overall survival rates comparable to stantially shorter time frame than standard IMRT treatments
surgical series when tumors receive a ­biological equivalent (27). Kilovoltage cone beam computed tomography (CBCT)
dose (BED10) greater than 100 Gy (1-4). can be used for image guidance during VMAT treatments
and has the potential to provide direct tumor tracking and
Severe toxicities (Grade 4-5) have been observed when adaptive radiation therapy solutions in the future (28-30). In
treating centrally located lung lesions with a dose of 60 Gy addition, respiratory-gated VMAT treatments are available
in 3 fractions (5-7). However, successful outcomes have for lung tumors, thus reducing the volume of normal lung
been reported for tumors in this location when increasing exposed to radiation (31).
the number of fractions or decreasing the dose per fraction
(8, 9). Currently, the Radiation Therapy Oncology Group The goal of this study was to compare the potential advan-
(RTOG) is conducting a dose escalation study (protocol tages of these three modalities for centrally located lung
0813) to identify safe fractionation schedules for this area lesions. Dosimetric parameters were compared as well as the
(10). Besides modifications of dose schemes, the implemen- risk of developing symptomatic radiation pneumonitis (RP).
tation of modern delivery systems may improve the thera- The risk of RP was calculated using normal tissue complica-
peutic ratio by means of superior dose distributions. Today, tion probability (NTCP) models.
the most advanced SBRT modalities include robotic radio-
surgery (Accuray Inc., Sunnyvale, CA), helical tomotherapy Materials and Methods
(Accuray Inc., Sunnyvale, CA), and volumetric modulated
arc therapy (Varian Medical Systems, Palo Alto, CA and Patient Selection and Contouring
­Elekta, Stockholm, Sweden)
Nine patients with centrally located early stage lung tumors
Robotic radiosurgery (RR) combines non-isocentric beam were selected according to the entry criteria for RTOG 0813.
delivery with image guidance and tracking techniques (11). Tumors were located within 2 cm of the proximal bronchial
Image guidance is achieved by registering a pair of orthog- tree and were 5 cm in size. All patients were treated with
onal x-ray images to digitally reconstructed radiographs RR SBRT between 2009 and 2012. Cases were classified
(DRRs) generated from the planning CT. Fiducial markers based on the PTV volume and on the proximity of the PTV
or anatomical landmarks are used to guide the image regis- to the adjacent critical structures. The latter quantity was
tration process. In addition, the tracking system can enable defined as the volume of overlap between the PTV and any
synchronization of respiratory-induced target motion with adjacent critical structure.
radiation delivery, thus allowing elimination of motion-
related tumor volume expansion (12, 13). Promising clini- For each patient, a plan for each modality (RR, VMAT,
cal outcomes have been reported in patients with peripheral HT) was created using identical planning scans and contour
lesions treated with RR (12, 14-18), but only limited studies sets for a prescription dose of 50 Gy in 5 fractions. Thus, a
are available for central lesions (8). total of 27 plans were generated. All patients had a whole-
thorax free-breathing CT scan at 1.5 mm slice thickness
Helical tomotherapy (HT) is a novel image-guided system and a 4DCT scan using a 16-slice CT scanner (Siemens
capable of delivering intensity-modulated radiation therapy Somatom Sensation). 4DCT image data were sorted into 8
treatments (IMRT) by combining a continuously rotating respiratory phases and maximum intensity projection (MIP)
fan beam with synchronous couch movement (19). Image images were generated. Tumor targets and critical structures
guidance is achieved by acquiring 3D images of the patient were contoured according to RTOG 0813 guidelines (10).
anatomy using the treatment (megavoltage) beam. Mega- The gross tumor volume (GTV) was delineated on the free
voltage CT (MVCT) imaging is highly integrated into the breathing scan with lung window settings, while the inter-
system and it is performed as part of every treatment fraction nal target volume (ITV) was delineated on the MIP dataset
to improve target localization and reduce setup uncertain- and then checked against scans from each respiratory phase.
ties (20-22). Previous studies have reported on the feasibil- The planning target volume (PTV) was created by expand-
ity and efficacy of using HT for hypofractionated treatments ing the ITV by a uniform 5 mm margin to compensate for
of centrally located early stage NSCLC and lung metastases setup error and residual respiratory motion not represented
(23-25). by 4DCT.

Technology in Cancer Research & Treatment 2013 December 6. Epub ahead of print
SBRT Dose Distributions in Central Lung 51

Planning and Beam Configuration PTV volume (TV). A value closer to unity represents bet-
ter target coverage.
For all treatment modalities, 6 MV photon beams were used.
RR and HT have unflattened photon beams, while for VMAT TC 5 TVPD/TV
the flattened photon beam was used. RR plans were gener-
ated using the Multiplan treatment planning system, version 2. Conformity index (CI): The CI is the ratio of the volume
4.5 (Accuray Inc., Sunnyvale, CA). The nominal dose rate receiving the prescription dose (VPD) to the PTV volume.
at isocenter was 1000 cGy/MU. Plans were optimized using The value should be ideally 1.2 or acceptably 1.5.
the iris variable aperture collimator (32) and the sequential
optimization method (33). The Monte Carlo dose calculation CI 5 VPD /TV
algorithm (1% uncertainty) was used for tissue heterogeneity
correction (34). 3. Conformation number (CN): The CN was proposed by
van’t Riet et al. (37). A value closer to unity represents
HT plans were generated using a binary multileaf collimator better target conformity.
(MLC) with a 6.25 mm projected leaf width at the isocenter
(which is 85 cm away from the photon source). Field widths CN 5 (TVPD /TV) 3 (TVPD /VPD)
of 1.05 or 2.5 cm and a pitch of 0.3 were used. The dose rate
at isocenter was 870 cGy/min. A modulation factor of 2 was 4. Heterogeneity index (HI): The HI is defined as the ratio
set at the beginning of the optimization process. The col- of the maximum dose to the prescription dose. A value
lapsed cone superposition-convolution algorithm was used closer to unity represents higher dose homogeneity and
for dose calculation (35). less heterogeneity.

VMAT plans were generated using the Eclipse treatment HI 5 Dmax /Dprescription
planning system, version 11, with 5 mm multileaf width and
Acuros calculation algorithm. The dose rate at isocenter was 5. Low dose fall-off
600 cGy/min. Combinations of two full arcs were used. The
analytical anisotropic algorithm (AAA) dose calculation • The gradient index (GI) is defined as the ratio of the vol-
algorithm was used for tissue heterogeneity corrections (36). ume receiving a given percentage of the prescription dose
to the PTV volume. Typically, GI is reported in terms of
Plan Evaluation 50% of prescription dose and is defined as by R50%. A
value closer to unity represents better dose fall-off.
A dose of 50 Gy in 5 fractions was prescribed. Acceptable treat-
ment planning met all dose specification criteria in RTOG 0813. R50% 5 V0.5 PD/TV

• Target coverage: 95% of the PTV is covered by the • D2 cm is defined as the maximum dose at 2 cm from
prescription dose and 99% of PTV is covered by 99% PTV in any direction.
of the prescription dose.
• Target dose heterogeneity: prescription isodose line is 6. Dose-volume parameters: The volume of lung receiving
between 60% and 90%. 5 Gy, 15 Gy, 20 Gy and 30 Gy (V5, V15, V20, V30) and the
• High dose spillage: any dose 105% of the prescrip- mean lung dose (MLD) for whole, ipsilateral, and contra-
tion dose (hot spot) is mainly located inside the PTV, lateral lung volumes were recorded. The lung volume was
and hot spots are not located inside any OAR (even if defined as the entire lung volume excluding the GTV. The
the OAR is part of the PTV). maximal dose was evaluated for other critical structures.
• Low dose spillage: the dose fall-off beyond the PTV is 7. Total monitor unit (MU) requirement and treatment time:
rapid and meets the criteria of RTOG 0813. These were based on the expected delivery parameters of
• Dose to critical structures (spinal cord, esophagus, bra- the machine at the time of treatment.
chial plexus, heart, trachea, proximal bronchial tree,
lung, skin and great vessels): maximum point dose and Risk of Radiation-induced Pneumonitis
dose volume limits are within the limits of RTOG 0813.
The risk of developing symptomatic radiation pneumoni-
The following parameters were collected and evaluated: tis (RP) strongly depends on lung dose-volume parameters.
To assess the probability of grade 2 (or higher) RP and to
1. Target coverage (TC): The TC is defined as the ratio of account for the biological effect of high fractional dose, an
target volume receiving prescription dose (TVPD) to the in-house Matlab program was developed.

Technology in Cancer Research & Treatment 2013 December 6. Epub ahead of print
 Kannarunimit et al.

The differential dose volume histogram (dDVH) of both be 0.43-0.45 for the whole lung (43, 44), and 0.67 for the uni-
lungs was re-scaled to the equivalent dose for 2 Gy/fraction lateral lung (45). The volume effect parameter, n, has been
using the Linear-Quadratic model (Eq. 1) with α/β 5 3 for estimated between 0.87-0.93 (44).
normal lung tissue (38):
We calculated the NTCP for bilateral and unilateral volumes
D /N separately. For the NTCP prediction using the bilateral lung
1 i
α/β volumes, we used TD50 5 20, n 5 0.87 and m 5 0.4. For
LQED2 i  Di , [1] the unilateral lung, we used m 5 0.67 and n 5 0.87. How-
2
1 ever, there is a wide range in TD50 estimates for the unilateral
α/β
lung. Based on the available estimates, we separately cre-
Here Di is the dose for bin i of the dDVH, N is the number of ated best (TD50 5 32.4), average (TD50 5 27.4), and worst
fractions, and α/β is the dose where the linear and quadratic (TD50 5 22.4) case scenarios to calculate the NTCP. NTCP
component of the survival curve are equal. curves were created by plotting the probability of RP as a
function of mean effective lung dose.
Normal tissue complication probability (NTCP) was derived
from Lyman-Kutcher-Burman’s (LKB) calculation model Statistical Analysis
(Eq. 2-4) (39-41).
All dosimetric parameters and toxicity probabilities were
1
t x2 summarized by the mean and standard deviation. The per-
NTCP  ∫ e dx, [2]
2 π ∞
2
cent of NTCP was plotted against the mean effective lung
dose using different settings of lung volume parameters for
Deff  TD 50 each technique (RR, HT and VMAT). To determine whether
t , [3] changes in the dosimetric parameters and NTCP were sig-
mTD 50 nificantly different among the modalities, a 1-way analysis

n
 of variance method for repeated measures was used. When
Deff  ∑ vi Di1/ n  , [4] the overall test indicated a statistical difference among the
 i  mean values, defined as a probability (p value) less than 0.05,
the post hoc Newman-Keuls test was used to identify which
Deff is the dose that, if given uniformly to the entire approaches differed from each other. There were no adjust-
­volume, will lead to the same NTCP as the actual non- ments for multiple comparisons.
uniform dose distribution. TD50 is the effective dose that
results in a 50% complication probability, m is a measure Results
of the slope of the sigmoid dose-response curve, and n is
the volume effect parameter. Note that this model is based The median age of patients was 60 years old (range 48-71).
on an effective dose (Eq. 4), similar to the generalized The majority of patients had T2 disease (67%) with average
equivalent uniform dose (gEUD), except the exponents GTV 11.6 6 7.86 cm3 and average PTV 36.8 6 18.1 cm3. The
are inverted. This assumes that the risk of complication median of the maximal tumor diameter was 3.28 6 0.7 cm.
is equivalent for the entire organ uniformly receiving the Four patients had PTV overlapping onto critical structures,
effective dose as the organ receiving an inhomogeneous
dose distribution. The effective dose is calculated from the Table I
volume (vi) and dose (Di) per bin in the LQED2-adjusted Patient and tumor characteristics (N 5 9).
dDVH obtained from Eq. 1.
Characteristic

The accuracy of NTCP prediction is based on the validity Median age in years (range) 60 (48-71)
of these 3 parameters (TD50, m and n for lung tissue). We T stage, N (%)
extensively reviewed these parameters to accurately estimate  T1 3 (33%)
 T2 9 (67%)
the risk of pneumonitis. Historically, these parameters for
GTV (cm3): Mean 6 SD 11.6 67.9
pneumonitis have been established from clinical data pertain-
PTV (cm3): Mean 6 SD 36.2 618.1
ing to conventional radiation treatment, normally defined as Median maximal diameter in cm (range) 3.3 (1.9-3.9)
TD50 5 24.5, m 5 0.18 and n 5 0.87 (42). More recent clini- Tumor location, N (%)
cal data from SBRT-related studies of pneumonitis show a  Anterior 1 (11%)
TD50 between 19.6-20.8 based on a normal tissue volume cal-  Middle 3 (33%)
 Posterior 5 (56%)
culated from the bilateral lungs (43, 44), and 22.4-32.4 Gy for
Median volume of PTV-organ overlap in cm3 (range) 0.02 (0-19.7)
the unilateral lung (45, 46). The value for m was reported to

Technology in Cancer Research & Treatment 2013 December 6. Epub ahead of print
SBRT Dose Distributions in Central Lung 53

Figure 1:  Dose distributions in the same axial plane for VMAT, HT, and RR plans (left to right). (A) Patient who had no PTV overlapping onto any critical
structures. VMAT, HT, and RR had equal index of target coverage and high dose conformity. All dose constraints on critical structure met RTOG criteria. RR
had better dose fall-off (R50% 5 4.2, D2 cm 5 19.9 while for HT and VMAT, R50% 5 5.3 and 4.6 and D2 cm 5 23.1 and 27.8, respectively). However, dose-
volume lung parameters and the risk of pneumonitis were similar between RR and VMAT, both better than HT (MLD 5 3.3, 3.9 and 2.9; V20 5 3.3, 4.5 and
3.8; V5 5 17.7, 18.5 and 11.4; risk of RP 5 3.4, 4.7 and 3.5% for RR, HT and VMAT, respectively). (B) Patient who had a large PTV without any overlap
onto critical structures. VMAT, HT, and RR had similar target coverage and high dose conformity. All dose constraints on critical structure met RTOG criteria.
RR had better dose fall-off (R50% 5 3.1, D2 cm 5 25.4 while for HT and VMAT, R50% 5 4.8 and 3.9 and D2 cm 5 31.6 and 32.4, respectively). However,
VMAT had better lung sparing than RR and HT (MLD 5 4.6, 5.6 and 5.4; V20 5 5.7, 5.5 and 7.4; V5 5 2.4, 3.2 and 3.2; risk of RP 5 7.6, 7.8 and 10.7% for
VMAT, CK and HT, respectively). (C) Patient who had the largest PTV with overlap onto the proximal airway and great vessel. VMAT, HT, and RR had
similar target coverage. RR and VMAT failed to limit the 105% hotspot in the region of overlap. RR also failed to limit R50% in this case. As a result, RR not
only decreased in conformity but also in low dose fall-off (CN 5 0.79, R50% 5 5.8 and D2 cm 5 32.9), while HT and VMAT maintained performance under
these conditions (CN 5 0.9 and 0.9, R50% 5 4.7 and 4.4, D2 cm 5 27.3 and 31.6, respectively).

Technology in Cancer Research & Treatment 2013 December 6. Epub ahead of print
 Kannarunimit et al.

for which the range of overlapping volume was 0.4-19.7 cm3 compared to HT and 91% compared to RR. VMAT reduced
(0.1, 0.2, 1.3 and 29% of PTV, respectively). Over half (56%) MU by 71% compared to HT and 80% compared to RR.
of the patients had tumors located in the posterior thorax
(behind the posterior border of the heart), 33% were in the Comparison of Other Organs at Risk
middle thorax (between the anterior and posterior border of
the heart), and 11% were in the anterior thorax. Patient and All plans except one achieved the RTOG 0813 dose con-
tumor characteristics are shown in Table I. straints for critical organs. As described above, in one case, RR
and VMAT failed to limit the 105% hotspot within the PTV
For all plans, the dose constraints of RTOG 0813 were overlapping on the great vessels. RR, HT and VMAT plans
achieved, except for one patient whose PTV was 67 cm3 resulted in similar doses to the proximal airway, heart, esopha-
with the largest area of overlap on the proximal airway and gus and spinal cord. Overall, RR plans delivered significantly
great vessels (19.7 cm3) (Figure 1C). For this patient, RR lower maximal doses to the great vessel, compared to VMAT
and VMAT failed to limit the 105% hotspot in the region (p 5 0.04). However, RR plans delivered significantly higher
of overlap. The maximal dose to the great vessels for this maximal dose to the brachial plexus, compared to both HT
case was 53.2, 52.5 and 53.8 Gy in RR, HT and VMAT, (p 5 0.0003) and VMAT (p 5 0.0002) as shown in Table III.
respectively. RR also failed to limit the ratio of R50% to be
less than 4.8 for this case. R50% was 5.8, 4.7 and 4.4 in RR, Comparison of Normal Lung Dose
HT and VMAT.
For the whole lung, RR and VMAT had a significantly lower
Comparison of Target Dose Distribution V5, V15 and V20 compared to HT as shown in Table IV.
VMAT had a significant lower D1000 and D1500 com-
All three techniques resulted in similar target coverage and pared to RR and HT (p 5 0.005 and p 5 0.006 for D1000,
dose conformity as shown in Table II. RR had a signifi- p 5 0.005 and p 5 0.02 for D1500). In addition, the lowest
cantly greater inhomogeneity than both of HT and VMAT MLD was obtained in VMAT as compared to RR and HT
(p 5 0.0002 for both comparisons). Regarding low dose (VMAT vs. RR p 5 0.005, VMAT vs. HT p 5 0.0002, RR
spillage and fall-off, RR and VMAT had a significantly bet- vs. VMAT p 5 0.02).
ter gradient index (R50%) than HT (p 5 0.002 and 0.04). RR
also had a significantly lower D2 cm compared to HT and For the contralateral lung, RR had significantly better contra-
VMAT (p 5 0.048 and 0.01). However, delivery efficiency lateral lung sparing compared to HT and VMAT as shown by
was superior for VMAT, HT and RR (p 5 0.0002 for all pair- lower V5 (p 5 0.0006 and p 5 0.0004) and lower contralat-
wise comparisons). VMAT reduced delivery time by 55% eral MLD (p 5 0.0002 and p 5 0.0008).

Table II
Characteristics of target dose distributions (Mean 6 SD).

RR HT VMAT p

Target coverage 0.96 6 0.01 0.96 6 0.01 0.96 6 0.01 0.64

Conformity index 1.11 6 0.09 1.10 6 0.05 1.11 6 0.05 0.88
Conformation number 0.84 6 0.06 0.84 6 0.05 0.84 6 0.03 1.00
Heterogeneity index 1.47 6 0.11 1.16 6 0.03 1.19 6 0.02 0.001
Gradient index: R50% 3.99 6 0.86 5.07 6 0.24 4.52 6 0.46 0.002
D2 cm 24.62 6 4.26 27.20 6 3.44 28.52 6 2.2 0.02
MU/fraction 13109 6 4165 8882 6 1217 2609 6 477 0.001
Minutes of time/fraction 54 6 12.88 10.35 6 1.45 4.66 6 0.78 0.001

Table III
Maximal dose delivered to nearby critical organs (Mean 6 SD).

  RR HT VMAT p

Proximal airway 41.19 6 11.65 43.26 6 10.41 39.52 6 11.63 0.53

Great vessel 38.46 6 17.31 39.81 6 14.92 41.52 6 15.96 0.01
Heart 25.24 6 12.71 23.60 6 12 23.27 6 14.82 0.53
Esophagus 18.98 6 13.38 18.09 6 10.44 20.63 6 11.15 0.13
Spinal cord 10.53 6 3.73 11.15 6 2.96 12.87 6 3.31 0.06
Brachial plexus 8.62 6 6.47 2.98 6 3.87 2.05 6 4.22 0.001

Technology in Cancer Research & Treatment 2013 December 6. Epub ahead of print
SBRT Dose Distributions in Central Lung 55

Table IV plans resulted in a significantly lower risk of symptomatic

Dosimetric parameters for whole and contralateral lung (Mean 6 SD). pneumonitis compared to HT.
  RR HT VMAT p
The risk of symptomatic RP was also predicted using uni-
Whole lung lateral lung volume parameters. We calculated unilateral
 V5 25.18 6 8.56 27.42 6 7.95 23.20 6 7.21 0.03 
 V15  6.92 6 3.27 10.96 6 2.76  7.69 6 1.62 0.001 
lung NTCP for three scenarios labeled as best (TD50 5 32.4),
 V20  4.46 6 2.11  5.66 6 1.54  4.66 6 1.26 0.002  average (TD50 5 27.4), and worst (TD50 5 22.4), using
 D1000  3.52 6 2.15  3.63 6 2.03  2.52 6 2.09  0.004 m 5 0.67 and n 5 0.87 as shown in Table VI and Figure 2.
 D1500  1.79 6 1.25  1.52 6 1.61  0.97 6 1.06  0.006 When using ipsilateral lung parameters, RR (p 5 0.004) and
 MLD  4.58 6 1.35  4.95 6 1.09  4.12 6 0.98  0.001 VMAT (p 5 0.0008) had a significantly lower risk of RP
Contralateral lung
compared to HT in all scenarios, as shown in Table VI. For
 V5 6.22 6 6.34 16.01 6 9.04 15.5 6 8.47 0.001 
 V10 0.76 6 1.41  0.94 6 1.63 1.06 6 1.74  0.86 contralateral lung parameters, RR plans had a significantly
 MLD 1.54 6 0.62  2.26 6 0.64 1.91 6 0.61 0.001  lower risk of RP compared to both HT (p 5 0.0002) and
VMAT (p 5 0.0003) in all scenarios.
Comparison of Probability of Symptomatic Radiation
Pneumonitis (NTCP) Table VI
Probabilities of symptomatic radiation pneumonitis by calculation parame-
ters by best, average, and worst case scenarios (mean% 6 SD).
The risk of symptomatic RP was estimated using the parame-
ters for a normal tissue volume of the whole lung (TD50 5 20, RR HT VMAT p
m 5 0.44, n 5 0.87). The NTCP results for each plan are Whole lung 6.2 6 4.1 7.6 6 3.6  5.8 6 2.6   0.02
shown in Table V. RR (p 5 0.04) and VMAT (p 5 0.02) Contralateral lung
 Average 7.6 6 0.5 8.3 6 0.5 8.1 6 0.5 0.001
Table V  Best 7.4 6 0.4 8.0 6 0.4 7.9 6 0.4 0.001
Individual case probabilities of symptomatic radiation pneumonitis  Worst 7.8 6 0.6 8.7 6 0.7 8.4 6 0.6 0.001
derived using whole lung volumes (%). Ipsilateral lung
 Average 22.0 6 6.5 24.4 6 7.0 21.1 6 5.9 0.001
Patient RR HT VMAT p
 Best 18.9 6 5.0 20.7 6 5.5 18.1 6 4.6 0.001
1 3.8  5.4  4.2    Worst 27.1 6 8.8 30.2 6 9.4 25.8 6 8.0 0.001
2 7.8 10.7  7.6  
3 15.9 12.3 10.2  
4 2.6  4.3  3.3  
5 4.7  5.3  4.5   Interestingly, for all scenarios (best, average, worst), the risk
6 3.4  4.7  3.5   of pneumonitis (NTCP) calculated from whole lung param-
7 8.5 13.3  9.1   eters was more consistent with the NTCP values estimated
8 3.8  4.9  3.8   from contralateral lung parameters rather than ipsilateral lung
9 5.7  7.4  6.1   parameters (Table VI). NTCP risk is about 2-3 times higher
Mean 6 SD 6.2 6 4.1 7.6 6 3.6 5.8 6 2.6 0.02
using only ipsilateral lung parameters. The contralateral or

Figure 2:  NTCP curves created by plotting the probability of symptomatic radiation pneumonitis as a function of mean effective lung dose for the unilateral
lung volume (left and center) and bilateral lung volume (right). For the unilateral lung model, curves for the three scenarios (best, average, worst) are presented.
Data for all the patients planned with the 3 modalities are shown for comparison.

Technology in Cancer Research & Treatment 2013 December 6. Epub ahead of print
56 Kannarunimit et al.

whole lung volume may be more useful in predicting RP Since all modalities achieve similar target coverage and
risk, particularly for arc-based radiotherapy such as HT and dose conformity, we propose that the relative benefits of
VMAT. Further studies are needed to validate this hypothesis. the techniques can be compared based on sparing of nearby
critical structures and reduced risks of toxicity. Our results
Discussion suggest that PTV size and degree of PTV overlap with cen-
tral structures (Figure 3) can be used to categorize patients.
Due to the proximity of adjacent critical structures, sophisti- For patients with no or small degree of overlap (10% of
cated radiotherapy techniques are advantageous when treat- the PTV volume), regardless of PTV size, RR provided the
ing centrally located lung tumors. In this study, we present best dose fall off and lowest lung dose. For patients with
the first direct comparison among three of these modalities: a large amount of overlap (10% of PTV), VMAT or HT
RR, HT and VMAT. The physical characteristics of the dose achieved good plan quality, while limiting hot spots on criti-
distributions, as well as predictive toxicity models, were used cal structures included within the treatment volume. How-
to guide the comparison. All plans were developed accord- ever, for patients with large PTV volumes, VMAT resulted
ing to the specifications of RTOG 0813, a current protocol in a potentially lower lung dose and lower risk of pneumo-
designed to identify safe treatment of centrally located lung nitis as well as decreased treatment time. These results were
lesions with hypofractionated dose regimens. obtained when using standard RTOG criteria to guide the
planning objectives. It is possible that adding additional plan-
Dosimetrically, all three modalities achieved similar target ning objectives focused on limiting normal lung dose might
coverage and conformity. While RR plans resulted in better alter the lung doses and risk of pneumonitis achieved by each
dose fall off at the edge of the target, VMAT and HT plans planning technique.
provided higher dose homogeneity and a significant reduction
in treatment time and monitor units. However, these results
were obtained using conventional full arc techniques for HT
and VMAT planning. It is possible that the addition of block-
ing parameters for HT planning or multiple partial arcs for
VMAT planning would result in sharper dose gradients at the
expense of delivery efficiency. In terms of normal structures,
RR, HT and VMAT achieved similar doses to proximal air-
way, heart, esophagus and spinal cord. RR plans provided a
lower maximal dose to the great vessels, compared to VMAT
and HT, but a higher dose to the brachial plexus.

We selected patients with a variety of tumor locations, dimen-

sions, and proximity to critical structures. To quantify the Figure 3:  Guideline for modality selection for SBRT for centrally located
lung tumors.
degree of “proximity” to OARs, we calculated the volume
of overlap between the PTV and the adjacent critical struc-
tures. One patient presented with a particularly large over- Previous authors have investigated planning and delivery
lapping volume involving both great vessels and proximal techniques for the treatment of lung cancer patients. How-
airways (19.7 cm3, equal to 29% of the PTV volume). For this ever, no study thus far has compared RR, HT and VMAT
patient, it was necessary to increase the dose homogeneity for centrally located lung lesions. Ding et al. (47) compared
within the target (the dose was prescribed to the 81% isodose RR with linac-based SBRT for eight lung cancer patients.
line), in order to push any dose 105% of the prescription They concluded that RR resulted in similar target cover-
dose away from OARs in the overlapping area. We observed age, less homogeneous dose to the target, and higher total
that in this situation, the RR performance was degraded MU. They also observed that RR plans resulted in a lower
not only in term of high dose conformity (as indicated by a dose to the lung for anteriorly located lesions and a higher
CN 5 0.79), but also in terms of low dose fall-off (as indi- dose to the lung for posteriorly located lesions. How-
cated by D2 cm 5 32.9 and R50% 5 5.8). On the other hand, ever, the patients in this study had peripheral lesions, and
HT and VMAT maintained these properties, while limiting hot dose calculations were not performed using Monte Carlo
spots in large overlap regions (for comparison, CN 5 0.9 and methodology.
0.86, D2 cm  5 27.3 and 31.6, and R50% 5 4.7 and 4.4, for HT
and VMAT). This observation seems to suggest that RR per- Meng et al. (48) compared HT and IMRT plans for ten lung
forms best when the dose homogeneity within the target is not cancer patients (5 central, 5 peripheral) undergoing conven-
constrained, whereas HT and VMAT can achieve good plan tionally fractionated radiation therapy. They concluded that
­quality despite the level of dose homogeneity within the target. HT provided better dose homogeneity, dose conformity, and

Technology in Cancer Research & Treatment 2013 December 6. Epub ahead of print
SBRT Dose Distributions in Central Lung 57

superior protection for OARs. However, HT plans resulted use whole lung volumes. Guckenberger et al. reported that
in larger volumes of lung receiving low levels of radiation, a grade 2 pneumonitis correlated with the mean lung dose to
tendency that was also found in our study. the ipsilateral lung volume (45). Ong et al. showed that con-
tralateral V5 significantly correlated to grade 2-3 pneumo-
Zhang et al. (49) compared plans generated with non-­coplanar nitis (55). In our study, the risk of symptomatic pneumonitis
3D, coplanar and non-coplanar VMAT, and flattening filter- was less than 10% when using whole lung and contralat-
free VMAT (FFF-VMAT) for fifteen lung cancer patients. eral lung volume parameters, but the risk was increased to
They concluded that VMAT plans not only shortened the 20-30% when using ipsilateral lung volume parameters. In
delivery time but also improved the plan quality in terms of the future, studies should report their findings with a care-
dose conformity to the target, dose fall-off in normal tissues ful annotation of the risk parameters being employed, given
and median dose to normal lung. that large differences can result depending on the method of
calculation.
Holt et al. (27) compared coplanar VMAT with coplanar and
non-coplanar IMRT for 27 early stage lung cancer patients Conclusion
with peripheral lesions eligible for SBRT. They concluded
that coplanar VMAT and non-coplanar IMRT achieved In this study, three SBRT techniques (RR, HT and VMAT)
similar plan quality, which was slightly better than coplanar for treating central lung lesions were investigated in terms
IMRT. However, VMAT significantly reduced treatment of their dosimetric characteristics, delivery efficiency, and
times, improving patient comfort. probability of radiation-induced symptomatic pneumonitis.
All techniques were able to provide clinically acceptable
Another major consideration in comparing different treat- plans following the guidelines provided by RTOG 0813.
ment modalities for lung SBRT is the technology-specific Target coverage and dose conformity were found to be
capability for online target localization. While image guid- very similar across modalities. Robotic radiosurgery and
ance is highly integrated in all three modalities, a unique VMAT resulted in a superior low dose gradient and lower
characteristic of the RR system is dynamic tumor tracking maximal dose to the great vessels. HT and VMAT resulted
(12, 13). In this study, the same contours were used for plan- in more efficient treatment delivery and in a higher level
ning all modalities, since our primary goal was to directly of target dose homogeneity. The estimated risk of RP for
compare physical characteristics of the dose distribution. each of the three modalities predicted that using whole lung
However, for patients treated with dynamic tracking, a tar- volume was less than 10%, similar to the risk found using
get volume without ITV expansion can be used, possibly contralateral lung volume. However, RP risk was 2-3 times
further reducing the risk of radiation-induced complications higher when calculated using ipsilateral lung volumes.
(50, 51). Using whole or ipsilateral lung volumes, RR and VMAT
had a lower risk of RP than HT, while for contralateral lung
One potential limitation to our study may be the use of the volumes, RR had the lowest risk of RP. Thus, RR produced
linear-quadratic (LQ) model to estimate RP risk. Park et al. a lower RP risk for a scenario of small PTV-OAR over-
(52) first proposed a better fit of the universal survival curve lap and small PTV. The greater homogeneity produced
(USC) model compared to LQ by evaluating the cell survival by HT and VMAT may be desirable for scenarios of large
curve from twelve NSCLC cell lines. However, while the ­PTV-OAR overlap. VMAT probably yields the lowest RP
USC might be reasonable to predict tumor control probability risk for a large PTV. Understanding subtle differences in
(TCP), its application to normal tissue complication probabil- the capabilities of these technologies may assist centers
ity (NTCP) remains unclear. Borst et al. (43) advocate using where multiple choices of modality are available. These
the LQ model for pneumonitis prediction, due to the potential investigations provide a basis for further investigations of
susceptibility of biological variables in the USC such as α, β, the modeling and normal lung parameters to be used in risk
Do, and Dq to unpredicted variations in heterogeneous cancer predictions related to SBRT.
tissue. We believe that the LQ model has sufficient practical-
ity for plan comparison, although further study is needed to Conflict of Interest
evaluate the validity of the LQ and USC models in toxicity
prediction. Dr. Descovich has received research support from Accuray, Inc.

The estimated symptomatic pneumonitis risk is 2-30% in Acknowledgments

published SBRT studies (44, 45, 53-59). One possible rea-
son for this variability is that some studies have considered The authors thank Accuray Incorporated and Varian Medical
the lung as a whole organ and did not distinguish between systems for providing access to a MultiPlan 4.5 workstation
the ipsilateral and contralateral lung. Two studies did not and Eclipse V.11 workstation, respectively.

Technology in Cancer Research & Treatment 2013 December 6. Epub ahead of print
 Kannarunimit et al.

References 11. Kilby, W., Dooley, J. R., Kuduvalli, G., Sayeh, S., Maurer, C. R.,
Jr. The CyberKnife Robotic Radiosurgery System in 2010. Tech-
1. Fakiris, A. J., McGarry, R. C., Yiannoutsos, C. T., Papiez, L., nology in Cancer Research & Treatment 9, 433-452 (2010). PMID:
­Williams, M., Henderson, M. A., Timmerman, R. Stereotactic body 20815415
radiation therapy for early-stage non-small-cell lung carcinoma: four- 12. Bibault, J. E., Prevost, B., Dansin, E., Mirabel, X., Lacornerie, T.,
year results of a prospective phase II study. International Journal Lartigau, E. Image-guided Robotic Stereotactic Radiation Therapy
of Radiation Oncology, Biology, Physics 75, 677-682 (2009). DOI: with fiducial-free tumor tracking for lung cancer. Radiat Oncol 7, 102
10.1016/j.ijrobp.2008.11.042 (2012). DOI: 10.1186/1748-717X-7-102
2. Martini, N., Bains, M. S., Burt, M. E., Zakowski, M. F., McCormack, 13. Hoogeman, M., Prevost, J. B., Nuyttens, J., Poll, J., Levendag, P.,
P., Rusch, V. W., Ginsberg, R. J. Incidence of local recurrence and Heijmen, B. Clinical accuracy of the respiratory tumor tracking sys-
second primary tumors in resected stage I lung cancer. The Journal of tem of the cyberknife: assessment by analysis of log files. Interna-
Thoracic and Cardiovascular Surgery 109, 120-129 (1995). PMID: tional Journal of Radiation Oncology, Biology, Physics 74, 297-303
7815787 (2009). DOI: 10.1016/j.ijrobp.2008.12.041
3. Onishi, H., Shirato, H., Nagata, Y., Hiraoka, M., Fujino, M., Gomi, 14. van der Voort van Zyp, N. C., Prevost, J. B., Hoogeman, M. S., Praag,
K., Niibe, Y., Karasawa, K., Hayakawa, K., Takai, Y., Kimura, T., J., van der Holt, B., Levendag, P. C., van Klaveren, R. J., Pattynama,
Takeda, A., Ouchi, A., Hareyama, M., Kokubo, M., Hara, R., Itami, P., Nuyttens, J. J. Stereotactic radiotherapy with real-time tumor
J., Yamada, K., Araki, T. Hypofractionated stereotactic radiotherapy tracking for non-small cell lung cancer: clinical outcome. Radio-
(HypoFXSRT) for stage I non-small cell lung cancer: updated results therapy and Oncology: Journal of the European Society for Thera-
of 257 patients in a Japanese multi-institutional study. Journal of peutic Radiology and Oncology 91, 296-300 (2009). DOI: 10.1016/
Thoracic Oncology: Official Publication of the International Associ- j.radonc.2009.02.011
ation for the Study of Lung Cancer 2, S94-100 (2007). DOI: 10.1097/ 15. Brown, W. T., Wu, X., Fayad, F., Fowler, J. F., Garcia, S., ­Monterroso,
JTO.0b013e318074de34 M. I., de la Zerda, A., Schwade, J. G. Application of robotic stereotac-
4. Uematsu, M., Shioda, A., Suda, A., Fukui, T., Ozeki, Y., Hama, Y., tic radiotherapy to peripheral stage I non-small cell lung cancer with
Wong, J. R., Kusano, S. Computed tomography-guided frameless ste- curative intent. Clin Oncol (R Coll Radiol) 21, 623-631 (2009). DOI:
reotactic radiotherapy for stage I non-small cell lung cancer: a 5-year 10.1016/j.clon.2009.06.006
experience. International Journal of Radiation Oncology, Biology, 16. Collins, B. T., Vahdat, S., Erickson, K., Collins, S. P., Suy, S., Yu,
Physics 51, 666-670 (2001). PMID: 11597807 X., Zhang, Y., Subramaniam, D., Reichner, C. A., Sarikaya, I.,
5. Le, Q. T., Loo, B. W., Ho, A., Cotrutz, C., Koong, A. C., Wakelee, H., Esposito, G., Yousefi, S., Jamis-Dow, C., Banovac, F., Anderson,
Kee, S. T., Constantinescu, D., Whyte, R. I., Donington, J. Results E. D. Radical cyberknife radiosurgery with tumor tracking: an effec-
of a phase I dose-escalation study using single-fraction stereotactic tive treatment for inoperable small peripheral stage I non-small cell
radiotherapy for lung tumors. Journal of Thoracic Oncology: Official lung cancer. Journal of Hematology & Oncology 2, 1 (2009). DOI:
Publication of the International Association for the Study of Lung 10.1186/1756-8722-2-1
Cancer 1, 802-809 (2006). PMID: 17409963 17. Vahdat, S., Oermann, E. K., Collins, S. P., Yu, X., Abedalthagafi, M.,
6. Timmerman, R., McGarry, R., Yiannoutsos, C., Papiez, L., Tudor, K., Debrito, P., Suy, S., Yousefi, S., Gutierrez, C. J., Chang, T., Banovac,
DeLuca, J., Ewing, M., Abdulrahman, R., DesRosiers, C., ­Williams, F., Anderson, E. D., Esposito, G., Collins, B. T. CyberKnife radiosur-
M., Fletcher, J. Excessive toxicity when treating central tumors in a gery for inoperable stage IA non-small cell lung cancer: 18F-fluoro-
phase II study of stereotactic body radiation therapy for medically deoxyglucose positron emission tomography/computed tomography
inoperable early-stage lung cancer. Journal of Clinical Oncology: serial tumor response assessment. Journal of Hematology & Oncol-
Official Journal of the American Society of Clinical Oncology 24, ogy 3, 6 (2010). DOI: 10.1186/1756-8722-3-6
4833-4839 (2006). DOI: 10.1200/JCO.2006.07.5937 18. Chen, V. J., Oermann, E., Vahdat, S., Rabin, J., Suy, S., Yu, X.,
7. Song, S. Y., Choi, W., Shin, S. S., Lee, S. W., Ahn, S. D., Kim, J. H., ­Collins, S. P., Subramaniam, D., Banovac, F., Anderson, E., Col-
Je, H. U., Park, C. I., Lee, J. S., Choi, E. K. Fractionated stereotactic lins, B. T. CyberKnife with tumor tracking: an effective treatment
body radiation therapy for medically inoperable stage I lung cancer for high-risk surgical patients with stage I non-small cell lung cancer.
adjacent to central large bronchus. Lung Cancer 66, 89-93 (2009). Frontiers in Oncology 2, 9 (2012). DOI: 10.3389/fonc.2012.00009
DOI: 10.1016/j.lungcan.2008.12.016 19. Kupelian, P., Langen, K. Helical tomotherapy: image-guided and
8. Nuyttens, J. J., van der Voort van Zyp, N. C., Praag, J., Aluwini, adaptive radiotherapy. Frontiers of Radiation Therapy and Oncology
S., van Klaveren, R. J., Verhoef, C., Pattynama, P. M., Hoogeman, 43, 165-180 (2011). DOI: 10.1159/000322420
M. S. Outcome of four-dimensional stereotactic radiotherapy for cen- 20. Ruchala, K. J., Olivera, G. H., Kapatoes, J. M., Schloesser, E. A.,
trally located lung tumors. Radiotherapy and Oncology: Journal of Reckwerdt, P. J., Mackie, T. R. Megavoltage CT image reconstruc-
the European Society for Therapeutic Radiology and Oncology 102, tion during tomotherapy treatments. Physics in Medicine and Biology
383-387 (2012). DOI: 10.1016/j.radonc.2011.12.023 45, 3545-3562 (2000). PMID: 11131183
9. Chi, A., Liao, Z., Nguyen, N. P., Xu, J., Stea, B., Komaki, R. ­Systemic 21. Ruchala, K. J., Olivera, G. H., Schloesser, E. A., Mackie, T. R. Mega-
review of the patterns of failure following stereotactic body radiation voltage CT on a tomotherapy system. Physics in Medicine and Biol-
therapy in early-stage non-small-cell lung cancer: clinical implica- ogy 44, 2597-2621 (1999). PMID: 10533931
tions. Radiotherapy and Oncology: Journal of the European Soci- 22. Welsh, J. S., Bradley, K., Ruchala, K. J., Mackie, T. R., Manon, R.,
ety for Therapeutic Radiology and Oncology 94, 1-11 (2010). DOI: Patel, R., Wiederholt, P., Lock, M., Hui, S., Mehta, M. P. Megavolt-
10.1016/j.radonc.2009.12.008 age computed tomography imaging: a potential tool to guide and
10. Bezjak, A., Bradley, J., Gaspar, L. E., Timmerman, R. D., Papiez, L., improve the delivery of thoracic radiation therapy. Clinical Lung
Gore, E., Kong, F.-M., Bae, K., Normolle, D. Seamless phase I/II study Cancer 5, 303-306 (2004). DOI: 10.3816/CLC.2004.n.010
of stereotactic lung radiotherapy (SBRT) for early stage, centrally 23. Chi, A., Jang, S. Y., Welsh, J. S., Nguyen, N. P., Ong, E., Gobar,
located, non-small cell lung cancer (NSCLC) in medically inoperable L., Komaki, R. Feasibility of helical tomotherapy in stereotactic
patients (RTOG 0813). Trial started February 2, 2009. In: ClinicalTri- body radiation therapy for centrally located early stage nonsmall-
als.gov [Internet]. Bethesda (MD): National Library of Medicine (US). cell lung cancer or lung metastases. International Journal of Radia-
(2009) [cited 2013 Nov 29]. Available from: http://clinicaltrials.gov/ tion Oncology, Biology, Physics 81, 856-862 (2011). DOI: 10.1016/
show/NCT00750269 NLM Identifier: NCT00750269 j.ijrobp.2010.11.051

Technology in Cancer Research & Treatment 2013 December 6. Epub ahead of print
SBRT Dose Distributions in Central Lung 59

24. Fuss, M., Shi, C., Papanikolaou, N. Tomotherapeutic stereotactic body of biologically related models for treatment planning: short report of
radiation therapy: Techniques and comparison between modalities. the TG-166 of the therapy physics committee of the AAPM. Medical
Acta Oncol 45, 953-960 (2006). DOI: 10.1080/02841860600897942 Physics 39, 1386-1409 (2012). DOI: 10.1118/1.3685447
25. Hodge, W., Tome, W. A., Jaradat, H. A., Orton, N. P., ­Khuntia, 39. Lyman, J. T. Complication probability as assessed from dose-volume
D., Traynor, A., Weigel, T., Mehta, M. P. Feasibility report of histograms. Radiat Res Suppl 8, S13-S19 (1985). PMID: 3867079
image guided stereotactic body radiotherapy (IG-SBRT) with 40. Kutcher, G. J., Burman, C. Calculation of complication probability
­tomotherapy for early stage medically inoperable lung cancer using factors for non-uniform normal tissue irradiation: The effective vol-
extreme hypofractionation. Acta Oncol 45, 890-896 (2006). DOI: ume method. International Journal of Radiation Oncology, Biology,
10.1080/02841860600907329 Physics 16, 1623-1630 (1989). PMID: 2722599
26. Rangaraj, D., Oddiraju, S., Sun, B., Santanam, L., Yang, D., Goddu, 41. Kutcher, G. J., Burman, C., Brewster, L., Goitein, M., Mohan, R.
S., Papiez, L. Fundamental properties of the delivery of volumetric ­Histogram reduction method for calculating complication probabili-
modulated arc therapy (VMAT) to static patient anatomy. Medical ties for three-dimensional treatment planning evaluations. Interna-
Physics 37, 4056-4067 (2010). PMID: 20879567 tional Journal of Radiation Oncology, Biology, Physics 21, 137-146
27. Holt, A., van Vliet-Vroegindeweij, C., Mans, A., Belderbos, J. S., (1991). PMID: 2032884
Damen, E. M. Volumetric-modulated arc therapy for stereotactic 42. Burman, C., Kutcher, G. J., Emami, B., Goitein, M. Fitting of normal
body radiotherapy of lung tumors: a comparison with intensity-mod- tissue tolerance data to an analytic function. International Journal
ulated radiotherapy techniques. International Journal of Radiation of Radiation Oncology, Biology, Physics 21, 123-135 (1991). PMID:
Oncology, Biology, Physics 81, 1560-1567 (2011). DOI: 10.1016/ 2032883
j.ijrobp.2010.09.014 43. Borst, G. R., Ishikawa, M., Nijkamp, J., Hauptmann, M., Shirato,
28. Lewis, J. H., Li, R., Watkins, W. T., Lawson, J. D., Segars, W. P., H., Bengua, G., Onimaru, R., de Josien Bois, A., Lebesque, J. V.,
Cervino, L. I., Song, W. Y., Jiang, S. B. Markerless lung tumor Sonke, J. J. Radiation pneumonitis after hypofractionated radiother-
tracking and trajectory reconstruction using rotational cone-beam apy: evaluation of the LQ(L) model and different dose parameters.
projections: a feasibility study. Physics in Medicine and Biology 55, International Journal of Radiation Oncology, Biology, Physics 77,
2505-2522 (2010). DOI: 10.1088/0031-9155/55/9/006 1596-1603 (2010). DOI: 10.1016/j.ijrobp.2009.10.015
29. Qian, J., Lee, L., Liu, W., Chu, K., Mok, E., Luxton, G., Le, Q. 44. Borst, G. R., Ishikawa, M., Nijkamp, J., Hauptmann, M., Shirato, H.,
T., Xing, L. Dose reconstruction for volumetric modulated arc Onimaru, R., van den Heuvel, M. M., Belderbos, J., Lebesque, J. V.,
therapy (VMAT) using cone-beam CT and dynamic log files. Sonke, J. J. Radiation pneumonitis in patients treated for malignant
Physics in Medicine and Biology 55, 3597-3610 (2010). DOI: pulmonary lesions with hypofractionated radiation therapy. Radio-
10.1088/0031-9155/55/13/002 therapy and Oncology: Journal of the European Society for Thera-
30. Kida, S., Saotome, N., Masutani, Y., Yamashita, H., Ohtomo, K., peutic Radiology and Oncology 91, 307-313 (2009). DOI: 10.1016/
Nakagawa, K., Sakumi, A., Haga, A. 4D-CBCT reconstruction j.radonc.2009.02.003
using MV portal imaging during volumetric modulated arc therapy. 45. Guckenberger, M., Baier, K., Polat, B., Richter, A., Krieger, T.,
Radiotherapy and Oncology: Journal of the European Society for Wilbert, J., Mueller, G., Flentje, M. Dose-response relationship for
Therapeutic Radiology and Oncology 100, 380-385 (2011). DOI: radiation-induced pneumonitis after pulmonary stereotactic body
10.1016/j.radonc.2011.08.047 radiotherapy. Radiotherapy and Oncology: Journal of the European
31. Qian, J., Xing, L., Liu, W., Luxton, G. Dose verification for respira- Society for Therapeutic Radiology and Oncology 97, 65-70 (2010).
tory-gated volumetric modulated arc therapy. Physics in Medicine and DOI: 10.1016/j.radonc.2010.04.027
Biology 56, 4827-4838 (2011). DOI: 10.1088/0031-9155/56/15/013 46. Ricardi, U., Filippi, A. R., Guarneri, A., Giglioli, F. R., Mantovani,
32. Echner, G. G., Kilby, W., Lee, M., Earnst, E., Sayeh, S., Schlaefer, A., C., Fiandra, C., Anglesio, S., Ragona, R. Dosimetric predictors of
Rhein, B., Dooley, J. R., Lang, C., Blanck, O., Lessard, E., Maurer, C. radiation-induced lung injury in stereotactic body radiation therapy.
R., Jr., Schlegel, W. The design, physical properties and clinical utility Acta Oncol 48, 571-577 (2009). DOI: 10.1080/02841860802520821
of an iris collimator for robotic radiosurgery. Physics in Medicine and 47. Ding, C., Chang, C. H., Haslam, J., Timmerman, R., Solberg, T. A
Biology 54, 5359-5380 (2009). DOI: 10.1088/0031-9155/54/18/001 dosimetric comparison of stereotactic body radiation therapy tech-
33. Schlaefer, A., Schweikard, A. Stepwise multi-criteria optimization niques for lung cancer: robotic versus conventional linac-based sys-
for robotic radiosurgery. Medical Physics 35, 2094-2103 (2008). tems. Journal of Applied Clinical Medical Physics/American College
PMID: 18561685 of Medical Physics 11, 3223 (2010). PMID: 20717090
34. Ma, C., Li, J., Deng, J., Fan, J. Investigation of fast Monte Carlo dose 48. Meng, L. L., Feng, L. C., Wang, Y. L., Dai, X. K., Xie, C. B.
calculation for CyberKnife SRS/SRT treatment planning. Med Phys ­Dosimetric comparison between helical tomotherapy and intensity-
34, 2589-2590 (2007). DOI: 10.1118/1.2761510 modulated radiation therapy plans for non-small cell lung cancer.
35. Ardu, V., Broggi, S., Cattaneo, G. M., Mangili, P., Calandrino, R. Chinese Medical Journal 124, 1667-1671 (2011). PMID: 21740774
Dosimetric accuracy of tomotherapy dose calculation in thorax 49. Zhang, G. G., Ku, L., Dilling, T. J., Stevens, C. W., Zhang, R. R.,
lesions. Radiat Oncol 6, 14 (2011). DOI: 10.1186/1748-717X-6-14 Li, W., Feygelman, V. Volumetric modulated arc planning for lung
36. Han, T., Followill, D., Mikell, J., Repchak, R., Molineu, A., Howell, stereotactic body radiotherapy using conventional and unflattened
R., Salehpour, M., Mourtada, F. Dosimetric impact of Acuros XB photon beams: a dosimetric comparison with 3D technique. Radiat
deterministic radiation transport algorithm for heterogeneous dose Oncol 6, 152 (2011). DOI: 10.1186/1748-717X-6-152
calculation in lung cancer. Medical Physics 40, 051710 (2013). DOI: 50. Nioutsikou, E., Seppenwoolde, Y., Symonds-Tayler, J. R., H ­ eijmen,
10.1118/1.4802216 B., Evans, P., Webb, S. Dosimetric investigation of lung tumor
37. van’t Riet, A., Mak, A. C., Moerland, M. A., Elders, L. H., van der motion compensation with a robotic respiratory tracking system: an
Zee, W. A conformation number to quantify the degree of confor- experimental study. Medical Physics 35, 1232-1240 (2008). PMID:
mality in brachytherapy and external beam irradiation: application to 18491515
the prostate. International Journal of Radiation Oncology, Biology, 51. Pepin, E. W., Wu, H., Zhang, Y., Lord, B. Correlation and prediction
Physics 37, 731-736 (1997). PMID: 9112473 uncertainties in the cyberknife synchrony respiratory tracking system.
38. Allen Li, X., Alber, M., Deasy, J. O., Jackson, A., Ken Jee, K. W., Medical Physics 38, 4036-4044 (2011). PMID: 21859002
Marks, L. B., Martel, M. K., Mayo, C., Moiseenko, V., Nahum, A. 52. Park, C., Papiez, L., Zhang, S., Story, M., Timmerman, R. D. Univer-
E., Niemierko, A., Semenenko, V. A., Yorke, E. D. The use and QA sal survival curve and single fraction equivalent dose: useful tools in

Technology in Cancer Research & Treatment 2013 December 6. Epub ahead of print
60 Kannarunimit et al.

understanding potency of ablative radiotherapy. International Jour- The British Journal of Radiology 85, 636-642 (2012). DOI: 10.1259/
nal of Radiation Oncology, Biology, Physics 70, 847-852 (2008). bjr/71635286
DOI: 10.1016/j.ijrobp.2007.10.059 57. Baumann, P., Nyman, J., Hoyer, M., Gagliardi, G., Lax, I., ­Wennberg,
53. Barriger, R. B., Forquer, J. A., Brabham, J. G., Andolino, D. L., B., Drugge, N., Ekberg, L., Friesland, S., Johansson, K. A., Lund, J. S.,
Shapiro, R. H., Henderson, M. A., Johnstone, P. A., Fakiris, A. J. Morhed, E., Nilsson, K., Levin, N., Paludan, M., Sederholm, C.,
A dose-volume analysis of radiation pneumonitis in non-small cell Traberg, A., Wittgren, L., Lewensohn, R. Stereotactic body radio-
lung cancer patients treated with stereotactic body radiation therapy. therapy for medically inoperable patients with stage I non-small
International Journal of Radiation Oncology, Biology, Physics 82, cell lung cancer - a first report of toxicity related to COPD/CVD
457-462 (2012). DOI: 10.1016/j.ijrobp.2010.08.056 in a non-randomized prospective phase II study. Radiotherapy
54. Matsuo, Y., Shibuya, K., Nakamura, M., Narabayashi, M., Sakanaka, and Oncology: Journal of the European Society for Therapeu-
K., Ueki, N., Miyagi, K., Norihisa, Y., Mizowaki, T., Nagata, Y., tic Radiology and Oncology 88, 359-367 (2008). DOI: 10.1016/j.
Hiraoka, M. Dose–volume metrics associated with radiation pneumo- radonc.2008.07.019
nitis after stereotactic body radiation therapy for lung cancer. Interna- 58. Nagata, Y., Takayama, K., Matsuo, Y., Norihisa, Y., Mizowaki, T.,
tional Journal of Radiation Oncology, Biology, Physics 83, e545-549 Sakamoto, T., Sakamoto, M., Mitsumori, M., Shibuya, K., Araki,
(2012). DOI: 10.1016/j.ijrobp.2012.01.018 N., Yano, S., Hiraoka, M. Clinical outcomes of a phase I/II study
55. Ong, C. L., Palma, D., Verbakel, W. F., Slotman, B. J., Senan, S. of 48 Gy of stereotactic body radiotherapy in 4 fractions for primary
Treatment of large stage I-II lung tumors using stereotactic body lung cancer using a stereotactic body frame. International Journal of
radiotherapy (SBRT): planning considerations and early toxicity. Radiation Oncology, Biology, Physics 63, 1427-1431 (2005). DOI:
Radiotherapy and Oncology: Journal of the European Society for 10.1016/j.ijrobp.2005.05.034
Therapeutic Radiology and Oncology 97, 431-436 (2010). DOI: 59. Timmerman, R., Paulus, R., Galvin, J., Michalski, J., Straube, W.,
10.1016/j.radonc.2010.10.003 Bradley, J., Fakiris, A., Bezjak, A., Videtic, G., Johnstone, D.,
56. Takeda, A., Ohashi, T., Kunieda, E., Sanuki, N., Enomoto, T., Takeda, Fowler, J., Gore, E., Choy, H. Stereotactic body radiation therapy for
T., Oku, Y., Shigematsu, N. Comparison of clinical, tumour-related inoperable early stage lung cancer. JAMA: the Journal of the Ameri-
and dosimetric factors in grade 0-1, grade 2 and grade 3 radiation can Medical Association 303, 1070-1076 (2010). DOI: 10.1001/
pneumonitis after stereotactic body radiotherapy for lung tumours. jama.2010.261

Received: September 7, 2013; Revised: October 10, 2013;

Accepted: October 11, 2013

Technology in Cancer Research & Treatment 2013 December 6. Epub ahead of print