Вы находитесь на странице: 1из 10

1

Ryan Salem

CSI Plan Challenge – Supine / Clinical Practicum III

Due: October 17, 2018

Supine, VMAT Technique for Craniospinal Irradiation

Patient Positioning, Setup, and Placement of Fields

For the Craniospinal Irradiation (CSI) Plan Challenge, I chose to complete treatment
planning on the supine dataset provided by ProKnow. The patient in this case was positioned
supine, with the arms at the sides. The knees are raised and the hands are at rest on the sides of
the body. A thermoplastic mask was which helps control head rotation and chin extension. The
chin is generally extended as much as possible without causing discomfort for the patient to
lower mandible dose.1

The dataset cut off the top of the head, so the fields extended from the top slice of the
dataset to the middle portion of the sacrum to encompass the entire planning target volume
(PTV), (Figure 1). Since volumetric modulated arc therapy (VMAT) with full arcs was used, all
anatomy and organs at risk (OAR) were within field borders and received some dose. Per the
ProKnow scoring metrics, the OAR where dose was limited included the lenses, optic nerves,
parotid and submandibular glands, thyroid, esophagus, lungs, heart, liver, kidneys, and bowel.
Other organs within the field included the eyes, brainstem, optic chiasm, mandible, trachea,
glottis, stomach, colon, sigmoid, and rectum. The femoral heads and bladder received low dose
as they were left below the level of the fields but were close to the most inferior field edge.

The plan was created on the Eclipse treatment planning system (TPS) version 13.6 for
treatment on a Varian TrueBeam accelerator. An SAD setup was chosen as the use of arcs was
used. I used 5 isocenters in this plan to make use of the finer leaves in the multileaf collimator in
the linac. The first isocenter was placed in the brain, and subsequent isocenters were made for
arcs to include the superior spine, middle spine, lower spine, and most inferior spine and sacrum
areas of the PTV. Isocenter was placed medially with only Y-directional (inferior/superior) shifts
between isocenters. Each subsequent isocenter from the brain was 15 cm inferior to the last
(Figure 1).
2

Figure 1. A sagittal slice showing the 5 isocenters used in this plan and the PTV_Total.

Treatment Planning and Plan Normalization

When compared to the typical 3D treatment of the whole central nervous system (CNS),
intensity modulation can be used to improve overall dose homogeneity. Although treatment
times are longer, McMahon et al.1 indicates that intensity modulated radiation therapy (IMRT)
plans can provide better target coverage and OAR sparing. Using a combination of research and
previously treated patients at my clinic by use of VMAT, I decided to create a plan using VMAT
arcs. Through the utilization of a single VMAT plan with multiple isocenters, areas of overlap
between fields are able to be modulated to avoid maximum dose and hot spots in these areas.1
Two full 360 degree arcs (180.1 – 179.9) were used at each isocenter with a beam energy of 6
MV. Collimator angles of 15 and 345 degrees were used for the brain fields, and angles of 5 and
355 degrees were used for the spine fields. No couch angles were utilized in this plan.

The plan was created in order to meet the planning metrics from the ProKnow scoring
sheet (Figure 2). The prescription for this plan was 36 Gy in 1.8 Gy fractions. The PTV total was
to receive 100% dose to 95% of the planning volume without going above 110%. Multiple
planning structures were used for treatment planning. First, a combined PTV_Total was created
for use in optimization and plan normalization. A planning PTV of the PTV_Total was made that
was cropped out of the optic nerves to help meet dose constraints. A PTV expansion by 1 mm in
all directions was added to help push dose into the PTV. Throughout various optimizations, the
PTV expansion was manually adjusted to allow for both pushing dose to areas not receiving
3

100% dose and also avoid areas receiving much dose. Small avoid structures were made in the
skull and brain to avoid maximum dose going above 39.6 Gy. A manual kidney avoid structure
was made to help meet the mean dose constraints of both kidneys (Figure 3a). A 2 mm lens
expansion was created to aid with dose fall around each lens (Figure 3b).

Figure 2. Scoring metrics for the Supine CSI plan challenge provided by ProKnow.

Figure 3a Figure 3b

Figure 3a The manually drawn kidney avoid (light green) can be seen in relation to both kidneys
(magenta and dark green) and the PTV (cyan) on axial and coronal views.

Figure 3b. The lens expansions are seen on axial and coronal views
4

The plan was optimized a number of times improve dose to the PTV and to address the
dose constraints listed above. A normal tissue objective (NTO) was used with priority of 150.
The NTO settings followed a 1 cm distance from target border, a start dose of 105%, end dose of
60%, and fall-off of .05. Structures such as the spinal canal and brain were changed to “control”
rather than “organ” so the optimizer did not try to limit dose to these structures. Through various
optimizations, avoid structures were added and the PTV expansion was edited to adjust dose
distribution. The metrics I had the hardest time meeting were consistently the mean dose to
kidneys, max dose to lenses, max dose to optic nerves, and the overall maximum dose limit of
39.6 Gy. To aid in meeting these constraints, I added skips to 3 of the 10 arcs. First, I 60 degree
skips to both of the brain arcs for each direction of 330-30 and 30-330 degrees. Doing this
allowed me to get both lenses below 7 Gy and both optic nerves below 34 Gy. I then added skips
to both lower spine fields which fully encompassed the bilateral kidneys. Isocenters were
intentionally placed so that the kidneys were completely inside one set of two fields and
completely out of the others. I added two 50 degree skips in each arc from 235-285 and 75-125,
respective to the direction. Doing this lost too much coverage of the PTV_Spine, so I then only
utilized the skips in the clockwise arc for the lower spine. This decreased my mean kidney doses
to below 2.5 Gy each, which came very close to meeting the 2 Gy constraint.

The plan was normalized so that 100% dose was delivered to 95% of my PTV_Total,
which was a combination of the PTV_Brain and PTV_Spine given to us in the dataset. Knowing
that I needed to achieve coverage to both of these, normalizing to the total ensured that 95% of
each volume was at least receiving 100% dose. Reference points were not used in this plan
because I was not particularly worried about the 100% isodose line reaching a certain area. Since
I was failing to meet multiple dose constraints, it was best to achieve minimum required target
coverage to keep OAR doses as low as possible. Hot spots were generally prevented by use of
avoid structures. It was easy to tell that the hot spots were a result of the heavy prioritization of
lens, optic nerve, and kidney avoidance in the brain and lower spine fields. Small areas of 39 Gy
were present but no point was over 39.6 Gy, or 110% prescription dose. The maximum dose
regions were in acceptable locations as they were found in the PTV_Brain with the maximum
dose point in the Brain_CTV. Cold spots were located in the same regions, again because of
avoidance. Below are images showing the dose distribution throughout the plan and the
cumulative DVH.
5

Figure 4. The 100% line around the PTV in the axial views of the PTV_Brian and PTV_Spine.

Figure 5. The 100% line in the sagital and coronal views in relation to the PTV_Total.
6

Figure 6. The 36 Gy and 2 Gy lines in the coronal view at the level of the kidneys.

Figure 7. The 36, 34, and 7 Gy isodose lines in the axial view at the level of the lenses and optic
nerves.
7

Figure 8. Axial view in dose color wash showing the target dose minimum of the PTV_Brain.

Figure 9. Axial view in dose color wash showing the target dose minimum of the PTV_Spine.
8

Figure 10. Cumulative DVH of the CSI plan with PTV, CTV, and OARs on the scoring sheet.

Plan Evaluation and ProKnow Scoring

My overall impression of the plan is that the VMAT technique with 5 isocenters was able
to produce a highly conformal plan capable of delivering 100% dose to 95% of the entire CSI
while sparing OARs. Due to the ProKnow limitations of the plan challenge, some clinically
significant sacrifices were made in order to score more points. This exact plan would not be used
clinically as the minimum point in the PTV_Brain is at 28.2% (Figures 8 & 10). In order to meet
the maximum point dose constraint of the optic nerves of 34 Gy, the PTV had to be cropped
9

significantly at the level of the optic nerves. Additionally, the kidney constraints were very tough
to meet as well, and PTV dose was sacrificed in order to achieve more points here. Additionally,
I would expect the mandible to be included in the OARs for actual patient treatment. The dose
constraints in my clinic for the optic nerves and kidneys is much more lenient, so I believe that
this technique would be extremely effective in CSI irradiation with modifications to improve
minimum target coverage. The use of 5 isocenters allowed for the finer MLC leaves in the
TrueBeam unit to block OAR structures and avoid maximum hot spots from being higher than
110%. In a clinical situation, it would probably be better to lower the number of isocenters to 3
or 4 to shorten overall treatment time for the patient, but still provide an acceptable treatment
plan.

This plan scored 126.02 / 127.00 possible points in the ProKnow system (Figure 11).
When converting the plan from eclipse to ProKnow, the optic nerve maximum dose was
different by over 1 cGy. This caused me to lose .49 points. Additionally, the PTV_Brain
coverage was slightly lost in the transition from the TPS to ProKnow, which also caused .16
points to be lost. The only constraint I failed to meet completely was the mean kidney constraint
of 2 Gy. I am okay with losing points here as I believe I pushed the system to its limit by
scripting so much dose to the PTV which was right next to both the kidneys.

Figure 11. Final ProKnow scoring.


10

Reflection

This plan challenge really pushed me pull out as many skills I have learned throughout
this year in order to achieve dose metrics. I created so many different structures and scripted
many different ways on the optimizer to finally achieve the plan I was looking for. This project
also served as a learning experience as well as I had to try many new techniques. I had to utilize
skip-arcs for the first time since I have been treatment planning. Additionally, I learned how to
use multiple isocenters in a single IMRT plan, something else I had not done before. I had to
push the limits of the planning computer with many avoid structures and strict constraints in the
optimizer, and further learned the capabilities of my clinic’s TPS.

I learned that this technique was used a few years ago in my clinic on a younger father of
2 daughters who was diagnosed with medulloblastoma in his early 40s. Five isocenters were
used and he was treated with a total CNS plan with a posterior fossa boost. The gentleman was
featured in a few magazines this year following his recurrence free recovery to this point, and he
has written thanks to the department for the care he received. The patient was on the table for
about 55 minutes each day for imaging and treatment of the primary CSI course.

Although old school and widely preferred 3D techniques are generally used in clinics
today, it was important to learn a different way to plan CSI radiation. If the patient can tolerate
the length of the treatment, it is important that this sort of treatment plan is offered if research
can back up the techniques. After trying to create a 3D plan for the supine dataset, I decided to
look into other methods as I was unable to score higher than 100 points on the metrics after quite
a bit of planning. Overall, this plan challenge forced me to learn 2 methods of CSI treatment, and
explore the advantages and disadvantages of both. I feel confident that if I am needed to create a
CSI plan in my future as a dosiemtrist, I will be able to create a high quality treatment plan for
my patient.

References:

1. McMahon RL, Larrier NA, Wu QJ. An image‐guided technique for planning and verification
of supine craniospinal irradiation. J Appl Clin Med Phys. 2011; 12(2): 184-190.
https://doi.org/10.1120/jacmp.v12i2.3310

Вам также может понравиться