Increased Local Control of Lung and Liver Tumors Associated with Dose-Escalated

1
Increased Local Control of Lung and Liver Tumors
Associated with Dose-Escalated Stereotactic Body
Radiation Therapy (SBRT) Supports a Dose-Response
Relationship Robert McCammon, MD, Tracey
Schefter, MD, Laurie Gaspar, MD, Rebekah
Zaemisch, MD, Daniel Gravdahl, MD, Brian
Kavanagh, M.D. Department of Radiation Oncology,
University of Colorado Cancer Center
Purpose/Objective To determine prognostic factors
for local control of primary or metastatic tumors
within the lung, liver, or other extracranial
sites treated with SBRT within a single
institution.
Results Local control was assessed
radiographically (example, Fig. 2). Dose was
considered either as nominal (prescribed) dose or
equivalent uniform dose (EUD), calculated from
the dose-volume histogram. On univariate analysis
(Log-rank), increased dose (nominal or EUD),
tumor within lung, and smaller GTV were
statistically significant predictors of increased
local control. On multivariate analysis (Cox
proportional hazards model), only dose (nominal
or EUD) retained significance (Table 1).
Patients/Methods The records of 152 consecutive
patients treated with SBRT from October 1999
through August 2005 were reviewed on an
IRB-approved protocol. A total of 261 lesions
in the lungs, liver, or other extracranial sites
were treated 18 (27/152) patients were treated
on prospective Phase I/II studies. Ineligible
patients or patients who declined to participate
in the dose-escalation studies were treated with
the highest safe dose established at the time of
treatment (Fig. 1). Nearly all patients (97)
were treated in a 3 fraction course.
Fig. 2. Characteristic fibrosis following
high-dose SBRT to right lung metastatic
lesion. Panel 1 Planning CT scan. Panels 2-6
3,6,12,18,24 month follow-up scans, respectively.
The patient has not recurred at time of analysis
(36 months image not shown).
Fig. 3. Actuarial local control according to
nominal dose. Lesions treated to a nominal dose
of gt54 Gy had a 3-year actuarial local control
rate of 89.3, whereas the values for those
treated to 36-53.9 Gy and lt36 Gy were 59.7 and
8.7 , respectively.
The local control rate of treated tumors
increased with escalated dose, expressed as
either nominal dose or EUD. (Figure 3). Likewise,
lesions treated with the highest third of EUD
(gt64.4 Gy) had a 3-year local control rate of
90.6, compared to 56.0 in the middle third
(43.5-64.4 Gy) and 9.9 in the lowest third
(lt43.5 Gy, plt0.0001).
Figure 4. The highest dose given to a lesion
near the proximal bronchial tree (see left 2
panels) was 54 Gy in 3 fractions. The patient did
not experience pulmonary toxicity and remained
locally controlled for 30 months before death
from intercurrent illness. However, the patient
did experience a Grade II vertebral fracture
attributed partly to SBRT, although documented
osteoporosis was a predisposing factor.
Treatment was well tolerated 3.4 of patients
experienced grade ?3 toxicity. Tumors near the
proximal bronchial tree were generally treated to
conservative doses (e.g. 36 Gy/3 fractions)
following reports suggesting unacceptable
toxicity associated with full-dose SBRT to that
site. An example of an unusual skeletal event in
such a case is shown in Figure 4.
Conclusions This large single institution series
suggests a dose-response relationship within the
range of SBRT doses applied. Severe toxicity was
uncommon. Excellent local control rates are
achieved with a nominal (prescription) dose gt 54
Gy in 3 fractions, corresponding to an EUD of
gt64.4 Gy. These results support the use of
aggressive, high-dose SBRT regimens when durable
tumor control is the primary objective.
Figure 1 Date of treatment versus total nominal
(prescribed) dose.