4

Principles of Radiotherapy
DAVID YOO x CHRISTOPHER WILLETT

KEY POINTS
The clinical application of radiation oncology balances
the dual goals of improved tumor control with reduced
treatment-related morbidity.
Over time, technological and biological advances have
offered many opportunities to further reduce normal
tissue complications and improve tumor control with
radiotherapy.

and how they may each in turn influence the decision-making

process.
This chapter will briefly describe these concepts, along with
specific clinical examples with various malignancies to illustrate
the interplay between the decision points. We will focus on the
therapeutic ratio, selection and definition of radiation target
volumes, side effects and toxicity considerations, dose fractionation, and the impact of technological advances on these topics.

More recent and emerging technologies promise further

improvements in the therapeutic ratio.

Therapeutic Ratio

Rational application and critical appraisal of these new

technologies will remain essential responsibilities for
the field of radiation oncology.

The overriding goal in patient care is to optimize treatment

efficacy while minimizing toxicity. This concept is represented
by the therapeutic ratio (Figure 4-2). Depending on the tumor
type and the surrounding normal tissues, a given amount of
radiation dose will yield an expected range of both tumor
control probability (TCP) and normal tissue complication
probability (NTCP). The vertical white lines illustrate how an
increase in dose (dashed line to solid line) leads to both higher
tumor control and normal tissue complications. An example
would be the dose escalation experience in prostate cancer.
Multiple randomized trials have shown an improvement in
prostate-specific antigen disease-free survival with higher
doses of radiation at the cost of increased rectal and urinary
complications.2
Ideally, the two curves can be further separated, with the
same radiation dose providing extremely high TCP and low
NTCP. Strategies that preferentially radiosensitize tumor cells
and shift the curve left (dashed blue curve) may allow for
improved TCP with the same physical dose. Alternatively,
techniques that preferentially protect normal tissues (dashed
red curve) may reduce treatment morbidity without compromising efficacy.
In clinical practice, agents that sensitize tumors often
affect normal tissues as well, leading to improved outcomes
at the cost of increased toxicity. Examples of this include the
use of concurrent chemotherapy with radiation compared to
radiation alone in the management of lung, head and neck,
cervical, esophageal, and gastrointestinal cancers.3-6 At the same
time, radioprotection agents have had less clinical impact than
expected. For example, the compound amifostine is U.S. Food
and Drug Administration (FDA)-approved for mitigation of
xerostomia in head and neck cancer patients treated with radiation alone.7 In practice, very few centers now utilize this agent
because of its cumbersome administration, its own toxicity
profile, the superiority demonstrated of concurrent chemoradiation in head and neck cancer, and the subsequent development of new radiation technologies that better spare the parotid glands.
One example where technological advances may have
improved the therapeutic ratio is the use of four-dimensional computed tomography simulation (4DCT) and intensity-modulated
radiotherapy (IMRT) planning in locally advanced lung cancers.8
39

Introduction
The discipline of radiation oncology involves the therapeutic
application of ionizing radiation in the management of benign
and malignant diseases. In fact, x-rays were used to treat a
patient with ulcerated breast cancer less than 2 months after
their discovery in November 1895.1 Therapeutic radiology, like
its diagnostic counterpart, has undergone considerable evolution since then. Still, the practice of radiation oncology has at
its core several fundamental concepts that have guided, and will
continue to guide, management decisions.
Tumors and their neighboring organs will remain in proximity to one another. These tumors and their normal tissue
counterparts will have varying responsiveness to ionizing radiation. Therefore, an ideal radiation treatment will balance optimal tumor control with acceptable acute side effects and late
toxicities. Over time, new technological or biological advances
will promise improved differentiation of these two populations,
with better clinical outcomes. Some of these advances will actually work, shifting paradigms and reinforcing the ongoing need
for rigorous testing before widespread adoption and changes to
accepted standards of care.
With each new patient encounter, the radiation oncologist
considers these concepts and determines the most appropriate treatment approach. A simplified algorithm is shown in
Figure 4-1, showing the general thought processes and decision points involved. The global decisions regarding treatment and goals of therapy are usually reached within a
multidisciplinary context with other medical providers and
informed consent of the patients and their support networks.
Once the decision has been made to incorporate radiation
therapy into the treatment plan, the more specific questions
regarding timing, dose, treatment volume, and technology
are considered. The solid and dashed arrows between the various boxes represent the interrelated nature of these questions

40

PART 1 Principles of Locoregional Therapy

TREAT?

Contraindications
Alternatives

Literature support

GOAL?

Cure
Palliation

Performance status
Patient preferences

Simulation with 4DCT allows for capture of respiratory motion to

avoid missing the tumor, whereas IMRT planning has the capacity
to deliver more conformal dose distributions with reduction in
high-dose normal tissue irradiation. In a retrospective series, the
use of 4DCT/IMRT compared with conventional computed tomography (CT) and planning resulted in improvements in overall
survival with decreased rates of radiation-induced pneumonitis.

Target Volumes
TIMING?

Neoadjuvant
Definitive
Adjuvant

Emergent
Urgent
Delayed

VOLUME?

Primary site
Nodal basins
Distant sites

Normal tissues
Critical organs
Margins

DOSE?

Fraction size
Total dose
Treatment time

Chemotherapy
Biologics
Hyperthermia

TECHNOLOGY?

Target delineation
Immobilization
Inverse planning

Image guidance
Adaptive therapy

100

100

75

75

50

50

25

25

Normal tissue complication probability %

Tumor control probability %

Figure 4-1 Treatment algorithm. This simplified algorithm shows the

general questions when encountering a patient under consideration for
radiation therapy. The questions are interrelated, as shown by the
dashed arrows to the right. Technological advances can often redefine
what is possible when it comes to treatment delivery. This, in turn, reemphasizes the importance of clinical investigations to demonstrate the
safety and superiority of these new techniques.

When radiation oncologists discuss treatment targets and volumes (Figure 4-3), they utilize definitions and acronyms developed by the International Commission of Radiation Units and
Measurements (ICRU). The ICRU 50 report published in 1993
introduced the concepts of gross tumor volume (GTV), clinical
target volume (CTV), and planning target volume (PTV).9 The
GTV represents grossly visible disease, either clinically or radiographically. The CTV represents areas of potential microscopic
extension of disease from the GTV. These microscopic deposits
may be in proximity to the primary tumor and/or dispersed
throughout the relevant draining regional lymph node basins.
The PTV includes sufficient margin around the CTV to ensure
adequate coverage and accounts for various uncertainties such as
patient/organ motion and daily setup errors.
Technologic advances such as 4DCT, which captures organ
motion, and immobilization devices/onboard imaging modalities that can reduce setup errors, have helped characterize and
eliminate some of these uncertainties. A modern linear accelerator treatment machine with onboard imaging and respiratory motion tracking capabilities is shown in Figure 4-4. The
subsequent ICRU 62 report considered these advances with the
addition of internal target volume (ITV) and setup margin
(SM).11 Now, PTV 5 CTV 1 ITV 1 SM. Further advances in
diagnostic radiology with functional imaging modalities are
providing potential biologic target volumes (BTVs) to further
improve the therapeutic ratio.12
These target volume definitions help to determine how much
technology and complexity is required to safely and effectively
deliver the radiation treatments to optimize the therapeutic
ratio. Depending on the clinical scenario, treatment plans may

PTV
ITV

Increasing dose

Figure 4-2 Therapeutic ratio. The blue solid curve represents tumor
control probability (TCP), whereas the red solid curve shows normal tissue
complication probability (NTCP). As radiation dose increases (dashed
vertical line to solid vertical line), both tumor control and toxicity are likely
to increase. The final dose chosen strikes the optimal balance between
the two outcomes. Various strategies can be employed to separate the
two curves. Factors that sensitize tumors to radiation may shift the TCP
curve to the left (dashed blue curve), whereas agents that protect normal
tissues may mitigate treatment-related toxicity, shifting the NTCP curve
to the right (dashed red curve).

Figure 4-3 Target volume definitions in radiotherapy. BTV, biological

target volume; GTV, gross tumor volume; CTV, clinical target volume;
ITV, internal target volume; PTV, planning target volume.

4 Principles of Radiotherapy

Motion detector
Gantry head
kV image
source

kV image detector

Treatment
couch
MV image detector

Figure 4-4 Linear accelerator treatment machine with onboard imaging and respiratory motion tracking devices. The radiation is delivered
via the gantry head that can rotate 360 degrees. The treatment couch
also rotates, allowing for various beam angle approaches. The kilovoltage imaging source and detector allow for acquisition of static portal
images as well as cone-beam CT images to confirm patient setup.
Megavoltage portal films may also be captured to ensure proper targeting. An infrared tracking device is seen in the ceiling, allowing for
detection of respiratory motion.

vary from relatively straightforward beam arrangements to

highly complex delivery methods with rigid immobilization and
onboard imaging modalities. For example, Figure 4-5 shows
representative treatment fields used to treat a patient with locally
advanced rectal cancer. Typically, the plan involves a posterior
anterior (PA) beam and two opposed lateral beams for coverage.
Tumor volumes and normal tissues are shown contoured on a
slice-by-slice basis and projected onto the digital reconstructed
radiographs. The primary tumor and nodal GTVs are shown in
red. The iliac blood vessels have been contoured in turquoise as
surrogates for sites of potential microscopic disease in the pelvic

41

lymph nodes. Small bowel (magenta) and femoral heads (tan

and blue) are also outlined. Panel A (PA) and panel B (left lateral
beam) demonstrate the additional margins on the GTV superiorly, inferiorly, and laterally to account for microscopic disease
in the mesorectum and pelvic nodes (CTV) and additional margin for setup uncertainties (PTV). Organ motion considerations
are often moot given the size of the treatment fields. Still, patients
undergo fluoroscopic evaluation with rectal contrast before
treatment to ensure the fields adequately cover the intended
volumes.
How do these target volumes fit into an algorithmic approach
for rectal cancer as described in Figure 4-1? When the decision
is made to treat a patient with locally advanced nonmetastatic
rectal cancer, the goal is cure and generally involves therapy
with surgery, radiation, and/or chemotherapy. Radiation therapy is often used in a neoadjuvant fashion, delivered preoperatively to the target volumes as outlined in Figure 4-5, prior
to surgical resection. The extensive literature on this subject
has described multiple fractionation regimens, including 5 Gray
(Gy) 3 5 daily treatments to a total dose of 25 Gy with radiation alone.13 A more typical strategy employed at Duke involves
1.8 Gy daily fractions to a total dose of 50.4 Gy with concurrent
chemotherapy.14 These volumes are large, with up to 2-cm margins
in all directions to account for any organ motion and setup
uncertainties.
The radiation doses are dictated by the tolerance of the small
intestines and by the fact that radiation therapy plays an adjuvant role to reduce locoregional recurrences following the primary treatment modality, which is surgical resection. Microleaf
collimators help to shield the small bowel and femoral heads.
Of note, the patient was also treated in the prone position
with a belly board designed to further minimize the amount of
small bowel irradiated by the lateral fields. Oftentimes, simple
maneuvers or changes in patient position during simulation
and immobilization can facilitate treatments without the need
for advanced planning and delivery techniques.

Figure 4-5 Three-dimensional treatment fields for rectal cancer. The beam apertures are custom-designed with multileaf collimators located in the
head of the treatment machine to encompass the target volumes with adequate margin while sparing surrounding normal tissue structures. Normal
tissues and tumor volumes are contoured and projected onto digital radiographic reconstructionleft femoral head (blue), right femoral head (tan),
small bowel (magenta), bladder (purple), rectum (yellow), iliac blood vessels (turquoise), and gross tumor volume (red). (A) Posterioranterior field.
(B) Left lateral field.

42

PART 1 Principles of Locoregional Therapy

Side Effects and Toxicity

Although some clinical scenarios may not benefit from the
utilization of advanced technological breakthroughs, the field
of radiation oncology remains dedicated to the rational application of these emerging technologies for any potential gains
in the therapeutic ratio. One such example may be found in the
evolution of adjuvant radiation therapy in the management of
breast cancer patients. Breast conservation with lumpectomy
followed by whole breast radiotherapy remains the current
standard, with radiation reducing the risk of recurrence by
two-thirds compared to surgery alone.15
Given the excellent survival and rates of cure, especially in
early-stage breast cancer patients, there is growing awareness
and emphasis on the potential late side effects of radiotherapy
that may reduce the therapeutic ratio.16 In particular, patients
with left-sided breast cancers often had anterior portions of the
heart within the tangential radiation treatment portals. This
excess heart irradiation likely contributed to the increased risk
of late cardiac morbidity and mortality seen in breast cancer
survivors.17 Subsequent techniques that tried to better shield
the heart may have compromised target coverage and resulted
in more recurrences under the heart blocks.18
With the advent of respiratory motion tracking, many leftsided breast patients are now simulated using breath-hold
techniques (Figure 4-6). Panel A shows the merged CT images
from a patient during both free-breathing and breath-hold at
end-inspiration. Panel B shows the medial border of a typical
tangential field for treatment of the whole breast. Portions of
the heart can be seen within the treatment volume. In panel C,

Breath-hold (BH)

Free-breathing (FB)

Heart
(BH)

the heart has physically moved away from the chest wall, allowing the same tangent field to cover the breast while minimizing
heart irradiation.
Because the late effects of radiation-induced changes to the
heart may require 1020 years to manifest, longer follow-up
and experience with this breath-hold technique are required
before the potential benefits of this change may be seen. Still,
the hope would be that the benefits of radiation already seen in
terms of locoregional control and survival would be further
enhanced if more patients were no longer dying from subsequent cardiac complications. Other strategies being employed
to minimize heart dose include treating patients in the prone
position or the use of IMRT to minimize heart dose in treatment planning. A much more radical change currently under
debate within the field is a redefinition of the appropriate CTV
itselffrom treating the whole breast to just the lumpectomy
cavity alone.19
Despite the continuing refinement and focus on target volumes, the tolerance of the surrounding normal tissues often
drives decisions regarding treatment. Acute side effects happen
during therapy and are often transient, requiring supportive care,
whereas late effects, as discussed, can develop months to years
following radiation therapy and tend to be irreversible. The issue
is further complicated by the fact that organs often have different
expressions of toxicity with different time courses. For example,
the skin may respond acutely to radiation with mild erythema
and dry desquamation. With increasing doses, patients may develop moist desquamation and even ulceration. Late effects may
include tanning changes, skin fibrosis and contracture, atrophy,
telangiectasias, or permanent hair loss.

Heart
(FB)

C
Figure 4-6 Utilization of breath-hold for treatment of left-sided breast cancer. (A) A merge of axial slices is seen from both free-breathing and deep
inspiration breath-hold CT simulation scans. The outer body contour shows increased chest wall expansion with deep inspiration. (B) Free-breathing
CT scan, with the thick yellow line representing the medial border of the tangential radiation fields used to treat the whole breast. In this scenario,
a small portion of the left ventricle is within the radiation treatment fields. (C) The use of deep inspiration breath-hold allows the same radiation
tangential fields to cover the left breast without irradiating the left ventricle.

4 Principles of Radiotherapy

Not all normal tissues are created equal. Some are organized
in a parallel fashion, with multiple independent functional subunits.20 For example, the kidney is composed of approximately
1 million separate nephrons, capable of performing their duties
independently from their neighboring subunits. Therefore,
organs arranged in parallel, such as the kidney, lung, and liver,
may have significant portions surgically removed or irradiated
beyond tolerance, with the remaining undamaged subunits able
to maintain normal global organ function. Other tissues, such
as the small bowel and spinal cord, are arranged in series, where
excessive damage along any point may compromise the entire
organs function.
These differences have significant implications in treatment
planning. When the target lies in proximity to a parallel structure, the dose parameters to avoid excessive toxicity are typically
expressed in mean dose or volume percentages. For example,
when evaluating lung cancer plans, a primary concern is the
potential development of radiation pneumonitis. One wellstudied dosimetric parameter is the volume of normal lung
receiving at least 20 Gy (V20), with higher percentages associated with increased risk of pneumonitis.21 For serial organs, the
goal is to avoid any dose beyond the expected tolerance of that
structure. With either the small bowel or spinal cord, the maximum dose is generally in the 4550-Gy range for conventionally
fractionated regimens.
Not all toxicities are created equal as well. Radiation-induced
cataracts may develop after doses beyond 10 Gy to the lens. These
can be easily removed surgically. Spinal cord myelopathy resulting in paralysis or blindness due to optic chiasm damage are
irreversible catastrophic complications. Figure 4-7 shows a patient
with nasopharyngeal carcinoma with intracranial extension
treated with concurrent radiation and chemotherapy.22 Panels A
and B demonstrate the utility of magnetic resonance imaging
(MRI)23 in the delineation of both tumor and normal tissues
compared to CT. The light-blue contour represents the 4400-cGy
isodose line. This dose level with chemotherapy is felt to be sufficient for microscopic disease control in the cervical neck nodes.
The yellow contour shows the 7000-cGy isodose line used for
gross disease control.
This is an example of an IMRT plan, where the intensity of
the beams is modulated during therapy to allow for more conformal dose distributions. The isodose lines show the radiation
dose being shaped away from the bilateral parotid glands and
the larynx in order to minimize long-term xerostomia and
edema/dysphagia. More critically, panel C shows the proximity
of the optic chiasm to the target volumes. Visual loss from radiation-induced optic neuropathy has been carefully studied and
reviewed. The risk is very low when the maximum dose to the
chiasm/optic nerves is less than 55 Gy at 1.82.0 Gy fraction
sizes. The incidence rises to 3%7% with maximum doses of
5560 Gy and rises quickly beyond 60 Gy.24 In this example, the
chiasm has been contoured and expanded by 1 mm to allow for
daily setup uncertainties (chiasm 1 1 mm). During IMRT
planning, priority was placed on covering the target volumes
while keeping the dose to the chiasm 1 1 mm structure less
than 54 Gy.
Plan evaluation is performed by careful examination of
the isodose distributions in the various axial, sagittal, and coronal planes and by review of a dose volume histogram (DVH)
that shows the various structures and their associated dosimetric characteristics (panel D). Based on the plan, the expanded
chiasm 1 1 mm structure is receiving a maximum dose of

43

5334 cGy. In essence, this 5334-cGy dose represents the worstcase scenario if the patients setup position is off by 1 mm every
day toward the high-dose areas. The chiasm itself is receiving
4731 cGy. Other organs, such as the parotids, are arranged in a
more parallel fashion. For these, the maximum dose has not
been found to be as critical as the mean dose received. Generally, mean doses less than 26 Gy are associated with improved
salivary function.25 Given the need to balance cure with toxicity,
and that patients are unlikely to die from xerostomia, target
volume coverage is usually given priority over parotid gland
sparing.
Over time, the typical tolerances of normal tissues to conventionally fractionated radiation have been meticulously gathered
and reexamined.26,27 The rapidly growing use of stereotactic techniques, which deliver much higher doses per fraction, has
spawned significant discussion and research in the establishment
of guidelines and tolerances of normal tissues to these larger
doses. For example, the recommended optic chiasm tolerance
for single fraction radiosurgery procedures is in the 1012-Gy
range.24

Dose Fractionation
In the earliest days of radiation therapy, treatments often involved
a single large fraction of radiation. Any clinical responses were
unfortunately accompanied by significant early and late toxicities.1 Subsequent strategies employed more protracted courses of
therapy, splitting the radiation treatment into multiple fractions
with smaller doses. Classical experiments from the 1920s with
rams testes showed the ability of small daily radiation fractions to
sterilize the animals without the resulting skin necrosis associated
with single large doses. The radiobiology behind fractionation
was best described by Withers and what he labeled the four Rs
repair of sublethal injury, cell repopulation, redistribution into
more radiosensitive phases of the cell cycle, and reoxygenation.28
Nonmalignant cells are more adept at the repair of radiationinduced DNA damage than their malignant counterparts. Therefore, the use of multiple radiation fractions allows for killing of
tumor cells, while giving the normal surrounding tissues time to
repair sublethal damage. One potential drawback of protracted
treatment courses is the opportunity for repopulation of both
normal and cancerous cells. For this reason, more fractionated
regimens typically require higher total doses to achieve similar
degrees of tumor control. Still, with each subsequent fraction, the
potential exists for more cancer cells to redistribute into the more
radiosensitive G2/mitosis phase of the cycle, allowing for more
effective radiation-induced cell kill. One final advantage of fractionation is the potential for reoxygenation of tumors during
treatment progression, as hypoxic cells are much more radioresistant compared to normoxic cells.
Different fractionation regimens have been used in a number of clinical scenarios. For example, common prescriptions
for the palliation of bone metastases include a single fraction of
8 Gy, 5 fractions of 4 Gy to a total dose of 20 Gy, 10 fractions of
3 Gy to 30 Gy, and 14 fractions of 2.5 Gy to 35 Gy.29 To make
comparisons between various schedules, radiation oncologists
often speak in terms of the biologically effective dose (BED) of
a particular treatment regimen. This equation takes into account a number of parameters, including total dose, fraction
size, overall treatment time, and the relative sensitivity of the
tissue to changes in fraction size.30 This final factor is known as
the alpha/beta ratio. Most tumors and normal tissues with

44

PART 1 Principles of Locoregional Therapy

PTV44

PTV70

chiasm
cord  5 mm

R parotid

chiasm  1 mm
L parotid

D
Figure 4-7 Intensity-modulated radiation therapy (IMRT) for nasopharyngeal carcinoma. Isodose distributions from the IMRT plan are superimposed
on coronal images from the CT (A) and MRI (B) simulation planning. The isodose lines show target volume coverage with sparing of the bilateral parotid
glands and larynx. The light-blue 4400-cGy isodose line covers all areas of potential microscopic tumor spread, whereas the yellow 7000-cGy isodose
line encompasses areas of gross disease. (C) The use of IMRT allows more precise coverage of target volumes while simultaneously sparing the optic
chiasm. To account for potential setup errors that could result in exceeding organ tolerance, the chiasm is expanded by an additional millimeter
(chiasm 1 1 mm) for treatment planning purposes. (D) The dose volume histogram (DVH) displays the dose delivered to various normal tissues and
target volumes.

4 Principles of Radiotherapy

rapid turnover (high alpha/beta ratios) are less impacted with

changes in fraction size. Some tumors and more slowly proliferating tissues (low alpha/beta ratios) are much more sensitive
to changes in fraction size, as seen with the optic chiasm.
To further illustrate this concept, consider a rapidly dividing
lung cancer (alpha/beta ratio 10) adjacent to the spinal cord
(alpha/beta ratio 2.5). Conventional 2-Gy fractions are used to
treat over 20 sessions to a total dose of 40 Gy. If the fraction size
is doubled to 4 Gy but the total dose is kept at 40 Gy, the tumor
will receive the equivalent of 47 Gy in 2-Gy fractions (17%
increase). However, the spinal cord will receive the equivalent of
58 Gy in 2-Gy fractions (44% increase). If the fraction size is
doubled again to 8 Gy, the tumor will receive the equivalent of
60 Gy in 2-Gy fractions (50% increase), whereas the spinal cord
will receive what feels like 93 Gy (133% increase). Because the
risk of myelopathy rises to unacceptably high levels beyond
50 Gy, regimens using higher doses per fraction will need to go
to a lower total dose that respects cord limits or employ techniques during patient simulation or treatment planning/delivery
to shield the spinal cord from irradiation beyond tolerance.

The Impact of Technological Advances

The ability to return to larger fraction sizes has been considerably aided by the various advances in target delineation, immobilization, organ motion capture, treatment planning, image
guidance, and verification. These advances can influence every
step along the decision-making process as initially shown in
Figure 4-1. As a case illustration, a 50-year-old woman presents
with a 1-cm brain metastasis in the left temporal lobe. Should
she be treated with radiation? Are there any contraindications
to radiotherapy? Is neurosurgical intervention an option? What
if she has newly diagnosed lung cancer with widespread systemic disease but without any neurologic sequelae? With no
neurologic symptoms requiring palliation, initiation of systemic
chemotherapy to address all her sites of disease may be the best
option, with radiation therapy to the brain reserved for symptomatic progression.
If she is treated with radiation, is the goal curative or palliative?
Historically, the development of brain metastases has represented incurable disease, with median survivals in the 27-month
range with radiotherapy.31 With improved systemic therapy and
identification of more favorable prognostic factors, further stratification has been performed, with median survivals approaching
21 months in young healthy breast cancer patients whose tumors
are positive for the hormone receptors and Her2-neu.32 Some
argue further for a paradigm shift, with more aggressive management of limited metastatic sites and the potential for cure in select
patients with oligometastatic disease.33 There is a growing body of
literature regarding aggressive treatment of metastatic disease,
initially with surgical resection and now with ablative radiation
therapy techniques.34,35
Over time, treatment strategies for patients with brain
metastases have changed in step with each technological advance.
Before the development of stereotactic radiosurgery, patients
with brain metastases not felt to be surgical candidates received
palliative whole brain radiation. A typical fractionation regimen
would be 10 daily fractions of 3 Gy to a total dose of 30 Gy over
2 weeks. With the advent of radiosurgery, institutions began to
investigate the potential benefits of boosting residual metastases
with large single fractions of 1524 Gy after whole brain radiotherapy. More recently, the long-term neurocognitive side effects

45

associated with whole brain treatment have led investigators

to study the use of radiosurgery alone to target individual
brain metastases. Despite multiple randomized studies and metaanalyses, the optimal treatment of brain metastases remains
controversial and continues to evolve.36
The decision to treat patients with whole brain radiotherapy
or radiosurgery (or both) has major implications for subsequent management choices. There are differences in treatment
volumes, total doses, fractionation schedules, and techniques
that necessitate varying degrees of patient immobilization,
simulation scans, the number of treatment portals, the resources required for treatment planning, quality assurance, and
delivery verification.
With whole brain radiation, a patient may be planned
clinically on the treatment machine itself without a separate
CT or MRI scan, immobilized with a piece of tape across the
forehead, with two opposing 17 3 23-cm portals and a custom
block placed in the field path to shield the eyes. The treatment
plan may be calculated by hand. Portal images may be taken
just prior to treatment to ensure adequate coverage of the intracranial contents. With a stereotactic radiosurgery case, the
patient requires rigid immobilization in a custom thermoplastic frame system, with fine-cut CT and MRI scans performed
and subsequently registered together. The lesion(s) and all normal tissues are contoured on a slice-by-slice basis and expanded
by a 12 mm margin to account for setup uncertainties. Various
plans with multiple beam arrangements and arcs are generated
and evaluated for adequate target coverage and respect for normal tissue tolerances. The plan then undergoes stringent quality
assurance. Just prior to treatment, the tumor location can be
more precisely matched with orthogonal films and cone-beam
CT images. An example of a stereotactic radiosurgery plan
is shown in Figure 4-8. Panel A shows the four arc paths chosen
to treat the lesion while minimizing dose to the brainstem
and optic structures. Panel B shows the arc paths with the field
apertures. The microleaf collimators move dynamically and
conform to the shape of the lesion as it changes along the arc
path. The resulting isodose distribution is shown, with a prescribed dose of 20 Gy.
As the field evolves and technology improves, institutions
are already studying the feasibility of delivering synchronous
simultaneous whole brain radiation with integrated boosts to
multiple metastases while minimizing dose to the hippocampus
to potentially reduce neurocognitive effects.37 Whether this
advance, or others to come, will lead to significant improvements in clinical outcome that justify their increased costs and
labor utilization is not known. As health care expenditures
come under further scrutiny, the field may find outside forces
ultimately determining what constitutes standard of care.
One final example of a technological advance changing the
practice of radiation oncology is that of stereotactic body radiation therapy (SBRT). For early-stage lung cancer, the standard
of care is surgical resection. Definitive radiation therapy has
been reserved for those patients unfit or unwilling to proceed
with surgery. Historically, the use of conventional radiotherapy
with daily 2-Gy fractions to a total dose of approximately 60 Gy
has been associated with poor local control and overall survival.38 Higher doses of radiation have been shown to consistently improve local control in lung cancer. With advances
in patient immobilization, respiratory motion capture, and
on-board imaging capabilities, SBRT has the ability to deliver
very high focal doses of radiation while minimizing irradiation

46

PART 1 Principles of Locoregional Therapy

C
Figure 4-8 Stereotactic radiosurgery for brain metastasis. (A) The target volume is shown in purple. The optic structures and brainstem
are also contoured. The fanlike structures show the paths of the four dynamic arcs used to deliver the radiation dose. (B) This reconstruction
shows the beams and field apertures. As the gantry arm moves along the arc path, the micro multileaf collimators continuously change
the field aperture shape to conform to the target volume. (C) The resulting isodose lines show the tight conformality of dose around the
metastasis.

of surrounding normal tissues. Various fractionation regimens

have been reported, from single doses of 34 Gy to five fractions
of 1012 Gy. This strategy has shown considerable improvements compared to historical series, with local control rates
higher than 90%.39 Overall survival numbers remain lower than
those reported for surgical series, but that is not unexpected,
given the competing risks in these inoperable patients. As the
field continues to evolve, studies are underway examining the
role of SBRT in operable lung cancer patients.
During simulation for an SBRT treatment plan, the patient
undergoes a 4DCT scan to assess the degree of organ and tumor motion. Depending on the size and location of tumors,
these lesions may or may not have significant excursion with
respiration (Figure 4-9). Panel A shows sagittal views from
each of the 10 phases of respiration captured by the 4DCT.
After evaluation of tumor motion, the decision is made
whether the patient is to be treated while breathing normally
or with breath-hold. Patients unable to hold their breath (or
unable to hear instructions) but with significant tumor motion

can still be treated using respiratory gating techniques, where

the radiation beam is turned on only during specific phases of
the respiratory cycle when the tumor is within the treatment
portal. Panel B shows a patient plan treated with free-breathing
technique. The tumor was contoured on both the free-breathing
images as well as the maximum intensity projection (MIP)
CT images. The MIP data reconstructs the imaging information from all 10 respiratory phase CT scans and represents the
degree of tumor motion due to respiration. This volume is
then expanded by only 5 mm to generate a PTV. Excess toxicity
has been reported with the use of 18 Gy 3 3 fractions for centrally located tumors near the proximal bronchial tree.40 As
a result, this patient was treated per the protocol from the
Radiation Therapy Oncology Group (RTOG) with 10-Gy
fractions 3 5 to a total dose of 50 Gy. Panel C shows the ninefield beam arrangement used to deliver the treatment. Just prior
to each treatment, a cone-beam CT scan is obtained on the
treatment machine to ensure the target volume is in right location (panel D). Finally, megavoltage cine images are obtained

4 Principles of Radiotherapy

47

E
Figure 4-9 Stereotactic body radiation therapy (SBRT) for early-stage lung cancer. (A) Sagittal images from 10-phase four-dimensional CT scan
(4DCT) demonstrate tumor motion during normal respiration. (B) Images from the4CT are used to generate a maximum-intensity projection (MIP) of
the tumor mass from all phases of the respiratory cycle. The isodose lines from the SBRT plan are superimposed, showing target volume coverage.
(C) The nine-field beam arrangement used for the treatment plan. (D) Cone-beam CT obtained on treatment table just prior to treatment to confirm
accurate target localization. (E) Cine-megavoltage images acquired during treatment delivery from one of the nine fields to confirm tumor coverage
throughout therapy.

during therapy to again confirm that the target volume remains

within the treatment portals (panel E).

Summary
Technological innovations continue to change the field of
radiation oncology. The dual goals of improved treatment efficacy and reduced toxicity remain at the core of the discipline.

Investigators will continue to explore the potential benefits that

come with each new generation of advances and ultimately how
they can improve patient care.
REFERENCES
The complete reference list is available online at www.
expertconsult.com.