Trial Information

Oropharyngeal Function After Radiotherapy With IMRT

A. Subjects One hundred and twenty-five patients with disease Stages III or IV squamous cell
cancers of the oral cavity, larynx or pharynx will serve as subjects.

B. Intensity Modulation and Radiotherapy Guidelines Anatomical data that is required for the
planning process is acquired using a Philip's AcQSim CT simulation system. The immobilized
patient is squared up relative to the bore of the Picker PQ2000S CT scanner using the
ceiling and side lasers of the system. A scout view is obtained so that the superior and
inferior borders of the volume study can be determined. Forty seconds prior to the
commencement of image data acquisition contrast media (125 cc of Ultravist 300) is power
injected at the rate of 1.3-2.0 cc/sec using a Liebel-Flarsheim device. The CT data set is
acquired with 3 mm slice thickness and spacing using the spiral mode of the scanner. While
the patient is on the table, the images are transferred to the VoxelQ work station that is
running AcQSim software version 4.1.1. Before the patient is released a reference isocenter
is established and recorded in the AcQSim system and marked on the face mask using the
isocenter locating lasers of the CT simulator, if not already indicated by radio-opaque
fiducials.

The physicians then outline target volumes, GTVs and CTVs, as per ICRU 50 guidelines and
outline critical structures such as the brain stem, spinal cord, salivary glands, optic
apparatus, etc. After this has been done, the image and structure data is transferred to
either of two treatment planning work stations: the NOMOS Corvus system running software
version 5.0 or the ADAC Pinnacle RTP system running software version 6.2b. A planning
target margin (PTV) is added to take into account organ motion and setup uncertainty.
Lateral and Anterior-Posterior digitally reconstructed radiographs (DRRs) that demonstrate
the location of isocenter are produced from the AcQSim system so that correct patient
positioning can be confirmed prior to the commencement of treatment.

The prescription for the target(s) defines the dose goal for the treatment, the minimal
acceptable dose and percent volume that can be under dosed and the maximum acceptable dose.
The prescription for radio sensitive structures sets the nominally allowed dose, the percent
volume that can be overdosed, the maximum tolerable dose and the minimum dose below which no
radiation damage can be observed. These specifications, for various targets and critical
structures, translate into defining three points on a dose volume histogram (DVH) curve for
each tissue type. The Corvus system computes the optimal beam intensity maps using a
simulated annealing algorithm whereas the ADAC system uses a sequential quadratic search
method to minimize a quadratic objective function, which is constructed from a set of
dose-based or dose-volume-based objectives for individual regions of interest. Delivery of
the dose as prescribed by the idealized maps is accomplished by using step and shoot IMRT at
both institutions. In this method the beam direction incident on the patient along with
field size and collimator angle is determined by the treatment planner before the software
designs the IMRT intensity maps. Therefore it is important to choose beam geometries
carefully so as to maximize exposure to the target and minimize exposure to critical
structures as visualized in the beam's eye view. The IMRT beams are created by utilizing a
standard 80 cm leaf MLC on an Elekta Sli linear accelerator.

Radiotherapy Guidelines for the proposed study:

1. All patients will have immobilization devices with treatment planning based on
Computerized Tomography (CT) information in treatment position. The CT slice thickness
through target will be 3mm.

2. CT and Magnetic Resonance Imaging (MRI) data will be used to define the various targets
and contour normal structures. MRI information will be used when indicated. Image
fusion will be used to relate CT/MRI data when necessary.

3. The field sizes and arrangements will be at the discretion of the attending Radiation
Oncologist. ICRU 50 guidelines will be used for various Tumor and Target Nomenclature:

1. The Gross Tumor Volume (GTV): All gross disease determined from clinical
examination and radiographic studies.

2. The Clinical Target Volume (CTV): The area that potentially contains microscopic
disease. Lymph node groups at risk of microscopic disease will be outlined as
part of corresponding CTV. The margin between each GTV and corresponding CTV can
vary from 0.2cm to 2.5cm depending upon the proximity to the critical and
uninvolved structures.

3. The Planning Target Volume (PTV): Provides a margin around CTV to compensate for
internal motion and set up errors. Typically a 5mm margin may be used around CTV.

Radiation Targets and Dose Specifications:

The following is the definition of targets. The radiation doses will be delivered as
specified in the protocol:

1. Primary Target Volume (PTV1): Includes gross tumor volume and volume encompassing
lymph nodes with moderate to high risk of metastases (first echelon nodes) and lymph
nodes with low risk of metastatic disease with corresponding CTVs.

2. Secondary Target Volume (PTV2): Includes gross tumor volume and volume encompassing
lymph nodes with moderate to high potential risk of metastases with corresponding CTVs.

3. Tertiary Target Volume (PTV3): Includes gross tumor volumes for the primary tumor and
nodal metastases with corresponding CTVs.

Low Neck: The midjugular, low jugular, and supraclavicular nodes can be treated either with
IMRT or alternatively with an AP field that is beam split to the IMRT fields. This will be
at the discretion of the treating physician. If an IMRT approach is not used, then it will
not be necessary to submit the DVHs for this PTV. However, if an IMRT approach is used,
then it will be necessary to generate DVHs for this PTV. Radiation doses will be as in the
attached protocol.

Neck Nodes: Guidelines for anatomic boundaries and CT-based delineation of neck nodes will
be as per consensus Atlas on RTOG website (www.RTOG.org → Researcher → H/N Atlas).

Normal Tissues: The appropriate normal organs will be contoured. Dose Volume Histograms
(DVH) will be generated. An attempt will be made to keep the maximum radiation doses to the
following organs as follows; since the radiation to these structures will be at a lower
fraction per day and the maximum doses to partial volumes only, we believe these are
acceptable dose levels.

Brain Stem: 54 Gy Optic Nerve/Chiasm: 55 Gy Spinal Cord: 45 Gy Mandible/TM Joint: 70
Gy or 1cc of the PTV not to exceed 75 Gy Larynx 70 Gy

Parotid Glands: Attempt will be to achieve mean dose 26 Gy in at least one gland or at
least 20cc of the combined volume of both parotid glands will receive 20 Gy or at least 50%
of the gland will receive 30 Gy (should be achieved in at least one gland).

Toxicity Reporting:

1. For acute toxicity the NCI/CTCAE Version 3.0 will be used.

2. For late toxicity, the RTOG/EORTC late toxicity criteria will be used.

Each treatment will be judged as being compliant with the Image guides therapy Center
in St. Louis guideline as:

1. No variation (Total Coverage): Coverage as specified in Radiation Guideline
Section.

2. Minor variation (Marginal Coverage): The 93% isodose covers between 95% to 98% of
the appropriate PTV or volumes of overdose exceed those specified in Radiation
Guideline Section by 5% of the PTV volume.

3. Major variation (Miss): The 93% isodose surface covers less than 95% of the
appropriate PTV or the overdose regions of the appropriate PTV are greater than 5% of
the PTV volume.

C. Data Collection Schedule Functional data will be collected on each patient at 6
points in time: pretreatment and 5 times after the completion of the chemoradiation --
1 month, 3 months, 6 months, 12 months, and 24 months.

1. Swallow Evaluation. At each evaluation point, each patient will receive a maximum
of 33 swallows studied by videofluoroscopy with concurrent manometry.

A 3 mm flexible catheter containing 2 solid state pressure sensors 3 cm apart will be
positioned transnasally at the beginning of the study, so that one sensor is located at
the base of the tongue and the second sensor is at the top of the upper esophageal
sphincter. This protocol will allow the measurement of pressures generated at the
tongue base. The concurrent manometry will allow us to directly measure pharyngeal
pressures during swallow. All swallows will be viewed in the lateral plane except the
last three 3 ml swallows viewed in the A-P plane. Subjects will not be required to
swallow volumes or viscosities that they cannot manage comfortably. All such
difficulties will be recorded on the data sheets.

Videofluoroscopy Protocol (3 swallows of each combination):

1 ml thin liquid barium 3 ml thin liquid barium 5 ml thin liquid barium 10 ml thin
liquid barium 3 ml pudding

¼ Lorna Doone cookie with 3 ml barium paste 3 ml thin liquid barium - chin down 3 ml
thin liquid barium - effortful swallow 3 ml thin liquid barium - super-supraglottic
swallow 3 ml thin liquid barium - Mendelsohn maneuver 3 ml thin liquid barium - P-A
view

2. Measure of Xerostomia. At each radiographic study of swallow, the patient's
stimulated saliva production will be quantified by taking the difference of the weight
of a folded sterile sponge (10 x 10 cm Kerlix) before and after chewing for 2 minutes.

3. Data Collection Regarding Onset and Nature of Oral Intake. At each of the 6
swallow data collection points, status and nature of oral and non-oral intake will also
be reported, as well as the reason for non-oral intake, i.e., nausea versus swallowing
disorder. The dates when the patient began to experience difficulty eating, when and
why non-oral feeding began, and when partial and full oral intake began will be
recorded and reported by the speech-language pathologist.

4. Ratings of Mucositis. Oral mucositis ratings will be completed pretreatment and at
the same 5 intervals after treatment. The scale rates ulceration on a 4-point scale (0
= no lesions to 3 = a lesion > 3 cm2) and erythema on a 3-point scale (0 = none; 1 =
not severe; 2 = severe) at 10 locations in the oral cavity (upper and lower lip, right
and left cheek, right and left ventral and lateral tongue, floor of mouth, soft
palate/fauces and hard palate). The mucositis ratings will be done by the
participating speech-language pathologist (SLP) trained in using the scale.

5. Patient's Perception of Function. A data form has been developed to define the
patient's perception of swallowing function, as well as dry mouth. The patient
completes the four-page form himself.

6. QOL scales: Performance Status Scale for Head and Neck Cancer (PSS-HN) (List et
al., 1990): This brief 3-part questionnaire will be given by the SLP at each
evaluation point; The M.D. Anderson Dysphagia Inventory (MDADI) is a 20-item
self-administered scale that investigates the impact of dysphagia on the patient's
quality of life; The University of Washington Quality of Life scale (UW QOL) is a 6
item clinician-administered scale.

7. Survival and Locoregional failure: Patients will be followed for overall survival,
relapse free survival and locoregional failure. The locoregional failure pattern will
be categorized as within a) Primary Target Region (PTV1); b) Secondary Target Region
(PTV2) or c) Tertiary Target Region (PTV3).