help me paper

International Journal of Radiation Oncology � Biology � PhysicsS114

258 High-Serum HGF and Low-Serum CD40L Are Associated With Liver Toxicity After Stereotactic Body Radiation Therapy K.C. Cuneo,1,2 M.A. Davis,2 M. Schipper,2 T.S. Lawrence,2 and M. Feng2; 1Ann Arbor Veterans Affairs Hospital, Ann Arbor, MI, 2University of

Michigan, Ann Arbor, MI

Purpose/Objective(s): We hypothesized that serum cytokine levels can predict radiation induced liver toxicity prior to clinical manifestations in

patients receiving liver SBRT.

Materials/Methods: As part of an IRB approved phase 2 trial of adaptive liver SBRT, patients were initially treated with 3 fractions of liver SBRT

and reassessed 1 month later. An additional 2 fractions of SBRT were

administered per protocol guidelines based on changes in indocyanine

green clearance. Blood samples were collected at baseline and 1 month

after the initial 3-fraction course of SBRT. Child-Turcotte-Pugh (CTP)

scores were prospectively acquired for each patient at baseline and 1, 3,

and 6 months after completing treatment. A 2 point increase in CTP score

was considered a clinically significant change. Serum cytokine levels were

analyzed using a bead-based multiplex assay. Fourteen cytokines were

selected based on their potential role in liver disease. Forty-five of the 90

patients on the study were selected for the initial screen, including all of

the patients who had a change in CTP score. Logistic regression models

were used to determine the probability of a 2 point change in CTP score as

a function of serum cytokine concentration. Receiver operating charac-

teristic (ROC) curves were generated and the area under the curve (AUC)

was calculated with confidence intervals.

Results: Baseline CTP scores were 5 or 6 in 34 patients, 7 or 8 in 10 patients, and 9 in 1 patient; 39 patients had hepatocellular carcinoma, 2

had cholangiocarcinoma, and 4 had liver metastases. Eleven patients

experienced a CTP score increase of 2 points or more within 6 months of

starting liver SBRT. Interestingly, the treatment planning normal tissue

complication probability for all of these patients was <1%. We found that

high-serum levels of hepatocyte growth factor (HGF) 1 month after 3 of

the 5 planned fractions of liver SBRT, was associated with an increased

risk of subsequently developing liver toxicity with an ROC AUC value of

0.83 (CI 0.70 to 0.97, P Z .003). Additionally, low-serum CD40 ligand obtained 1 month after 3 fractions of SBRT was associated with an

increased risk of developing liver toxicity with an ROC AUC value of 0.85

(CI 0.69 to 1.00, P Z .007). Conclusion: Serum levels of HGF and CD40L may be early biomarkers of radiation induced liver toxicity from SBRT. High-serum HGF has previ-

ously been shown to be elevated in patients with acute and chronic liver

disease. The low levels of serum CD40L seen in our study may be related

to liver sequestration of activated T cells targeting the epithelium. These

findings need to be validated in a larger patient cohort but could aid in

personalized adaptive radiation therapy.

Author Disclosure: K.C. Cuneo: Research Grant; GI SPORE. Head of

department; Ann Arbor Veterans Affairs Hospital. M.A. Davis: None. M.

Schipper: None. T.S. Lawrence: Research Grant; NIH. Partnership; PI

Squared Therapeutics. M. Feng: Panel member; NCCN.

259 Continuous Positive Airway Pressure (CPAP) for Motion Management in Stereotactic Body Radiation Therapy (SBRT) to the Lung: A Controlled Pilot Study J.D. Goldstein,

1 Y. Lawrence,

1,2 S. Appel,

1 T. Rabin,

1 M. Ben David,

1

N. Weizman,1 M. Benayun,1 S. Dubinski,1 D. Alezra,1 H. Gnessin,1

R. Levitin,2 M.J. Segel,1 N. Peled,1 and Z. Symon1,2; 1Sheba Medical

Center, Ramat Gan, Israel, 2 Sackler School of Medicine, Tel Aviv

University, Tel Aviv, Israel

Purpose/Objective(s): Continuous positive airway pressure (CPAP) has long been used in patients with obstructive sleep apnea to maintain airway

patency. There are no reports describing CPAP use with radiation therapy.

We performed a prospective clinical trial to evaluate the effects of CPAP

on tumor motion, lung volume, and dose to critical organs in patients

receiving SBRT for lung tumors.

Materials/Methods: Institutional review board (IRB) approval was given in December 2013. Inclusion criteria included adult patients with primary

or secondary lung tumors referred for SBRT. All patients underwent 4D

CT simulation twice: once with free breathing and then with CPAP. Tumor

motion was assessed by comparison of the internal target volume (ITV).

Volumetric and dosimetric parameters were compared using a Student’s

paired, two-tailed t-test; CPAP was utilized during treatment if judged

beneficial.

Results: Continuous positive airway pressure was tolerated well in 10 out of 11 patients enrolled. One patient declined treatment due to mask discomfort.

Ten patients with 18 lesions were evaluated. The use of CPAP significantly

decreased the amplitude of tumor motion with a mean ITV reduction of 27%

(95% CI: 16%e37%, P < .001). Lung volume was increased with use of

CPAP by 32% (95% CI: 21%e42%, P < .001), contributing to a 22% reduction in mean lung dose (95% CI Z 13%e32%, P < .001). Significant reduction (P <.01) was seen in lung: V5:17% (95% CI Z 7:28); V10: 31% (95% CI Z 19:42) and V20:24% (95% CI Z 6:43). The use of CPAP was also associated with a reduction in mean heart dose of 29% (95% CI Z 23%e36%, P <.001). Significant reduction (P <.01) was also seen in heart

V5: 32% (95% CI Z 23:42) and V10: 41% (95% CI Z 22:60). Ten patients (91%) were treated and completed all fractions using CPAP without acute

complications. One patient had grade 2 pneumonitis.

Conclusion: In this pilot study, CPAP used during lung SBRT was safe, well tolerated, and provided clinical and dosimetric benefit in almost all

patients. Compared to free breathing, the use of CPAP significantly

reduced lung tumor motion. The smaller ITV, planning target volume

(PTV), and the increase in total lung volume associated with CPAP

contributed to a reduction in lung and heart dose; CPAP was well tolerated,

reproducible, and simple to implement in the treatment room. Clinical

trials are in progress studying additional applications and optimal methods

for use of this novel technique.

Author Disclosure: J.D. Goldstein: Patent/License Fees/Copyright; Sheba

Medical Center, Physician, joint patent application re: use of CPAP in

radiation therapy. Y. Lawrence: None. S. Appel: None. T. Rabin: None.

M. Ben David: None. N. Weizman: None. M. Benayun: None. S.

Dubinski: None. D. Alezra: None. H. Gnessin: None. R. Levitin: None.

M. J. Segel: None. N. Peled: Patent/License Fees/Copyright; Sheba

medical center, joint patent application for use of CPAP in radiation

therapy. Z. Symon: Patent/License Fees/Copyright; Sheba medical center,

joint patent application for use of CPAP in radiation therapy.

260 Percussion Assisted Radiation Therapy (PART): A Unique Innovative, Noninvasive Approach to Control Respiratory Motion, Improve Imaging, and Radiation Therapy for Lung Cancer E.M. Ozsahin,1 N. Peguret,2 M. Zeverino,1 B. Belmondo,1 F. Duclos,1

J. Simons,1 O. Long,1 J.O. Prior,1 R. Moeckli,2 and J. Bourhis1; 1Centre

Hospitalier Universitaire Vaudois (CHUV), Lausanne, Switzerland, 2 Lausanne University Hospital, Lausanne, Switzerland

Purpose/Objective(s): Pulmonary tumors are subject to respiratory mo- tion inducing imaging artefacts, and often, require large volume treatments

by radiation therapy (RT). Gating techniques can solve these issues by

stabilizing lung targets and sustaining breath-holds in maximal inspiration

(MI); however, these are limited by the patient’s capacity to hold his/her

breath. The purpose of this study was to implement a new high frequency

percussive ventilation assistance method (PART: percussion-assisted RT),

and report its first clinical use in maintaining breath holds long enough to

optimize chest imaging and complex RT treatments.

Materials/Methods: After evaluating its tolerability in a cohort of 10 vol- unteers, ethical approval was obtained to conduct a pilot feasibility study

into clinical practice. We applied this novel PART technique successfully in

patients eligible for lung stereotactic RT (SBRT), locally advanced lung RT,

or left breast cancer RT. Dosimetric parameters in free breathing (FB),

maximal inspiration (MI) gating, and PART conditions were compared.

Volume 93 � Number 3S � Supplement 2015 Oral Scientific Sessions S115

Results: Percussion-assisted RT offered a high level of success in chest stabilization with excellent tolerability. The mean duration time for

breathing stabilization was 10.4 minutes in 10 healthy volunteers, 28 tests

(each subject tested 3 times), and 380 measures. We had 100% success on

first and second tests, and 80% success on the third tests. Transferred into

clinic, PART was applied with good tolerance for patients included so far,

without treatment breaks during the overall fractionated RT; PART offered

better dosimetric profiles when compared to MI or FB. For lung cancer

patients, the volume of normal lung tissue irradiated by the 95% isodose was

divided by a factor of two in the first two cases. Until now, we treated one

patient with left breast cancer (50 Gy in 25 fractions, 100% success rate,

mean stabilization for beam-on time 5.68 min, range 3.12e7.18 min), one patient treated with lung SBRT (60 Gy in 8 fractions, 100% success rate,

mean stabilization for beam-on time 10.00 min, range 8.55e12.68 min), and

one patient with locally advanced lung cancer (51 Gy in 17 fractions, 100%

success rate, mean stabilization for beam-on time 9.35 min, range

8.17e11.58 min). In addition to this, percussive ventilation assistance

markedly improved FDG PETimage quality in detecting pulmonary lesions.

Conclusion: Percussion-assisted RT was tolerable and applicable in clin- ical practice with a dosimetric gain for patients in comparison to other

breath hold gating techniques. This pilot study is ongoing, and updated

data will be presented at the meeting.

Author Disclosure: E. Ozsahin: None. N. Peguret: None. M. Zeverino:

None. B. Belmondo: None. F. Duclos: None. J. Simons: None. O. Long:

None. J.O. Prior: None. R. Moeckli: None. J. Bourhis: None.

261 Implementation of Real-Time, Real-Anatomy Tracking and Radiation Beam Control on the First MR-IGRT Clinical System O.L. Green, L.J. Rankine, B. Cai, R. Kashani, L. Santanam, S.M. Goddu,

C.G. Robinson, P.J. Parikh, J.R. Olsen, J.D. Bradley, and S. Mutic;

Washington University School of Medicine, St. Louis, MO

Purpose/Objective(s): We reported on the geometric and dosimetric re- sults, quality assurance, and workflow process for the first clinical

implementation of real-time, real-anatomy tracking and beam control for

free-breathing patients.

Materials/Methods: The first clinically implemented magnetic resonance imaging guided radiation therapy (MR-IGRT) system used for over a year at

our institution has the capability for sagittal cine imaging at 4 frames per

second, while tracking a chosen anatomical organ at the same time. The

tracking algorithm implemented in the delivery system deformed an initial

contour on every subsequent cine frame and compared it to a predetermined

boundary contour. If the anatomy of interest was determined via the

deformed contour to be outside the boundary, the radiation delivery was

stopped until it returned to within the boundary. To understand the system

performance and derive appropriate treatment margins and gating bound-

aries, the following components were characterized: overall system latency,

shutter dose, spatial integrity of the magnetic field, output constancy, and

absolute spatial dosimetric accuracy. Motion phantoms compatible with the

magnetic field were used with radiographic film to establish the spatial ac-

curacy of the gated dose distributions; MR-compatible ionization chambers

were used to confirm an implemented shutter dose correction. The overall

goal of these tests was to determine parameters such that the PTV margins

remained comparable to Linac-based deliveries, while delivering doses to

targets visualized within treatment boundaries.

Results: Gated output measurements showed a 1.28% average difference as compared to static delivery, demonstrating that the shutter dose correction

was appropriate. The spatial distribution of dose was within 2 mm of the

planned distribution when delivered to a motion phantom with the tracking

capability implemented. The overall latency of the system (500 msec on

average) contributedthe most to the spatial dose error. The spatial integrity of

the magnetic field was within 2 mm for a 24 cm diameter field of view, and 1

mm at 10. Coupled with inherent imaging resolution, it was determinedthat a

3 mm boundary around the anatomy of interest allowed for a 5 mm GTV to

PTV expansion, identical to standard of care for fiducial-based tracking in

our clinic. The expected duty cycles will be about 50% for the recommended

parameters, used with an exhale-based volumetric MR image for initial

setup. Five tracking-enabled deliveries with larger PTVexpansions showed

good tracking and beam control so far at duty cycles of 80e90%.

Conclusion: The goal of clinical real-time anatomy tracking has now been achieved in radiation therapy. By characterizing system performance in

relation to adjustable parameters, accurate dose distributions were achieved

while having ongoing monitoring of anatomy position during delivery.

Author Disclosure: O.L. Green: Honoraria; ViewRay, Inc. L.J. Rankine:

None. B. Cai: None. R. Kashani: None. L. Santanam: None. S. Goddu:

None. C.G. Robinson: None. P.J. Parikh: Research Grant; Varian, Phi-

lips. J.R. Olsen: Research Grant; ViewRay, Inc. J.D. Bradley: Research

Grant; ViewRay, Inc. S. Mutic: Honoraria; ViewRay.

262 Early Experience in Cone Beam Projection Image Streaming for Real-Time Intrafractional Motion Monitoring Using a Conventional Linear Accelerator Y.K. Park,1 G.C. Sharp,1 S.J. Ye,2 and B. Winey1; 1Massachusetts General

Hospital, Boston, MA, 2 Seoul National University Hospital, Seoul, Korea,

Republic of Korea

Purpose/Objective(s): Kilovoltage x-ray image streaming capability has been recently designed and enabled for linear accelerators, which poten-

tially enables real-time tumor localization and tracking on conventional

linear accelerators. This study aimed to investigate the feasibility and

performance of real-time fiducial tracking during cone beam CT acquisi-

tion with or without simultaneous treatment delivery.

Materials/Methods: A client computer was connected to an imaging system via a Gigabit Ethernet switch. The imaging system streamed out projection

image data to the client computer immediately after image acquisition with

fluoroscopy or CBCT. A performance test was conducted to investigate

imaging frequency, latency and tracking feasibility. Three gold fiducial

markers with a length of 3 mm were implanted in solid water slabs that

constituted a pelvis-shaped phantom with a thickness of 36 cm. The phantom

was placed on a motion platform oscillating with amplitudes of 1.4 cm and 3

cm in anterior-posterior and superior-inferior direction simultaneously, at a

cycle of 4 s. In-house marker tracking software based on FFT normalized

cross-correlation was developed and installed in the client computer. The

marker motion was tracked in real-time for two different imaging scenarios:

(1) a kV-only CBCT scan with treatment beam off; (2) a kV CBCT scan

during a 6 MV VMAT delivery. In total, 345 and 154 consecutive pro-

jections, respectively, from kV-only and kV-VMAT scans were analyzed.

Results: The projection streaming system successfully transferred projec- tion images to the client computer with a frequency of 5.8 +/- 1.8 Hz in kV-

only mode. In the kV-VMAT case, on the other hand, the frequency varied

according to the MV beam-on state. The system latency was found to be

231.2 +/- 9.8ms, including panel read-out time (w182ms). In addition, w150ms was required for image processing and marker detection algo- rithms. Using an in-house fiducial marker tracking algorithm, the detection

rate was >95% in the kV-only scan and 85.1% in the kV-VMAT scan.

Conclusion: Our early experience in projection streaming service demonstrated promising results supporting the feasibility of real-time

intrafractional motion monitoring using a conventional linear accelerator.

Further research is ongoing to improve tracking accuracy, MV-scatter

suppression and processing speed.

Author Disclosure: Y. Park: Research Grant; Elekta. G.C. Sharp:

Research Grant; Elekta. S. Ye: None. B. Winey: Research Grant; Elekta.

263 Interfractional and Immobilization Related Respiratory Motion Variability for Treatment of Liver Tumors B. Lavajo Vieira,1,2 T.K. Kosak,2 Y. Belkacemi,1,3 J.Y. Wo,2 A.X. Zhu,2

T.S. Hong,2 and J.A. Wolfgang2; 1(AROME) Association of Radiotherapy

and Oncology of the Mediterranean Area, Paris, France, 2 Massachusetts

General Hospital, Harvard Medical School, Boston, MA, 3 Henri Mondor

University Hospital, Créteil, France