Trial Information

A Phase II Study of Erlotinib (Tarceva) and Hypofractionated Thoracic Radiotherapy for Patients With Advanced or Inoperable Non-Small-Cell Lung Cancer

Hypothesis/Rationale: Lung cancer is the number one cause of cancer-related mortality in
both men and women in the United States, with over 170,000 cases diagnosed annually. The
overall 5-year survival rate remains 14% despite decades of clinical research. Thoracic RT
is the standard treatment for locally advanced (Stage III) NSCLC, in combination with
chemotherapy in favorable patients. Metastatic lung cancer (Stage IV) is treated with
systemic chemotherapy, with the addition of RT for palliation of tumor symptoms. Most lung
cancers present as large tumors, measuring 2 to 7 cm in largest dimension. It is therefore
not difficult to understand that only 16% of patients experience a complete resolution of
their irradiated tumors within 3 months following a course of standard fractionated (2.0 Gy
daily) RT and chemotherapy.

From basic principles advocated by Fletcher, it is thought that standard fractionated RT
doses up to 100.0 Gy may be necessary to sterilize tumors of the size frequently encountered
in clinical practice. Tumor control probability for bronchogenic carcinoma can be estimated
to be 10% for tumors of greater than 4 cm at a dose of 80.0 Gy, with an assumption, that an
average-size lung cancer may require doses beyond 100.0 Gy standard fractionated to have a
50% to 80% probability of controlling the tumor. This has been demonstrated in the
"stereotactic radioablation" approach to patients with the medically inoperable Stage I
NSCLC, in whom 20 Gy per fraction to a total of 60 Gy (a BED equivalent of >100 Gy) resulted
in an excellent local control of >90% (Timmerman R et al, 2006; Onishi H et al, 2007).

Hypofractionated RT: It is generally accepted that the presence of chronically hypoxic cells
within tumors may be an important cause of radioresistance and resultant local failure in
radiotherapy, particularly in large solid tumors such as locoregionally advanced NSCLC.
In-vitro experiments indicated that the dose needed to kill severely hypoxic cells is on the
order of 2 or 3 times the dose needed for oxic cells. Therefore, delivering a higher RT dose
to the tumor may result in higher tumor cell kill and improved local control. One of the
approaches to increase RT dose is to use hypofractionated RT, which not only increases the
dose, but also reduces the overall treatment time. The radiobiological rationale for
hypofractionated RT was described by Mehta et al (Mehta et al, 2001). Based on these
theoretical assumptions, University of Wisconsin has recently completed a dose escalation
study of progressively increasing fraction sizes in thoracic RT for lung cancer. Such
larger RT fraction sizes may require "tighter" radiation fields (to achieve reliable
normal-tissue sparing) and improved precision of RT delivery, something that is afforded by
the SBF (Stereotactic Body Frame) immobilization and daily CT scan based image verification
of tumor position. More experiences have been reported in the literature on
hypofractionated regimens for lung cancer. Japanese investigators (Nagata Y et al, 2002)
treated 40 patients with T1-T3N0 tumors or lung metastases with 10-12 Gy per fraction to a
total of 40-48 Gy. No pulmonary (or other) complications >Grade 2 were observed and the
local control was 100% in the subgroup of primary lung tumors. Another group from Japan
(Onimaru R et al, 2003) reported on 45 patients with primary lung tumors up to 6cm receiving
7.5 Gy per fraction to 60 Gy (lesions <3cm) or 6 Gy per fractions to 48 Gy (lesions 3-6cm).
One patient with a central tumor died of a radiation-induced ulcer in the esophagus. One
patient with a peripheral lesion experienced Grade 2 chest wall pain. The 3-year local
control rate was 80%. No adverse respiratory events were noted.

The "ultimate" hypofractionated RT regimen of 60 Gy given in 3 fractions of 20 Gy each has
been demonstrated to be feasible and highly effective in patients with medically inoperable
Stage I NSCLC tumors measuring up to 7 cm and located outside the central airways (Timmerman
et al, 2006). However, the same regimen was associated with a high incidence of severe
toxicity if applied to central airway. Since most tumors in patients with Stage III and IV
NSCLC are located centrally, a novel hypofractionated regimen needs to be developed
specifically for them.

In our institution we have completed a Phase I/II investigator-initiated trial of
dose-escalated hypofractionated RT given concurrently with Gefitinib (Iressa) with Gefitinib
continued after RT completion until progression or toxicity. Three RT dose levels are
applied: 4.2 Gy in 10 fractions to 42 Gy; 4.2 Gy in 12 fractions 50.4 Gy and 4.2 Gy in 15
fractions to 63 Gy. Eligible pts are those with either Stage III or IV NSCLC who needed
thoracic RT and could not receive chemotherapy. No selection by the EGFR receptor status
has been applied. A total of 12 patients have been enrolled. Main toxicities were
pulmonary (1 grade 2 pneumonitis; 1 grade 3 infectious pneumonia; 1 grade 4 pneumonitis).
There was 1 grade 3 abdominal pain. One patient (with thoracic tumor controlled) died due
to the late radiation-esophageal toxicity (tracheo-esophageal fistula in a setting of
pre-existing esophageal diverticula) at 12 months from RT. Only one patient experienced
local progression of the irradiated tumor, which is very encouraging and may support the
hypothesis of the radiosensitizing effect of the EGFR inhibitors. As of now, median
survival time for all 12 enrolled patients is 9 months (range: 1-26 mo) from the time of
initiating Gefitinib, which is an encouraging result in mostly pretreated patients, many
with metastatic disease (Werner-Wasik et al, oral presentation, First ESMO/IASLC European
Meeting on Lung Cancer, Geneva, Switzerland, April 2008).

Since Gefitinib is not available now for wider use, we are proposing a study of erlotinib
with hypofractionated RT in a Phase II setting with the main objective of assessing efficacy
of such a combination. There is no direct evidence that patients receiving concurrent EGFR
inhibitors and RT need to be pre-selected with regard to the EGFR status. In a recent study
(Bentzen SM et al, 2005), positive immunohistochemical staining for EGFR status was
associated with a benefit in locoregional control in patients with head and neck cancer
receiving CHART (continuous hyperfractionated accelerated radiotherapy), but the EGFR status
had no effect on survival or rate of distant metastases. Therefore, we propose to
investigate the EGFR status in all eligible patients, but to treat them without selection
for the EGFR status.

Tarceva (previously known as OSI-774) is an orally active, potent, selective inhibitor of
the EGFR tyrosine kinase. Early clinical data with Tarceva indicate that the compound is
generally safe and well tolerated at doses that provide the targeted effective concentration
based on nonclinical experiments. A recently completed, randomized, double-blind,
placebo-controlled trial (BR.21) has shown that Tarceva as a single agent significantly
improves the survival of patients with incurable Stage IIIb/IV NSCLC who have failed
standard therapy for advanced or metastatic disease (Shepherd F et al, 2000 and 2005). In a
Phase II clinical trial (Jackman D et al, 2007) of 80 chemotherapy-naive patients >70 years
of age with advanced non-small cell lung cancer who received erlotinib as first-line
therapy, an encouraging median survival time (MST) of 10.9 months was reported, with the
presence of EGFR mutations strongly correlated with disease control and survival.

In summary, a combination of erlotinib with hypofractionated thoracic RT has the potential
to significantly improve local tumor control in patients with non-small-cell lung cancer,
based on theoretical considerations of EGFR inhibition, increased tumor cell killing with
larger RT fractions and preclinical evidence for synergism between RT and erlotinib.

Our hypothesis is that the addition of erlotinib to RT will result in radiosensitization,
therefore increasing the likelihood of local tumor control over RT alone. Maintenance
erlotinib upon RT completion will result in further tumor growth inhibition, both
systemically and locally, lengthening disease-free survival and overall survival.

All eligible patients will be enrolled, without regard for the EGFR status. The implications
of prospectively screening patients for EGFR mutations or gene copy number and how the
patients should be selected for subsequent treatment remain to be defined (Janne P et al,
2005; Shepherd f et al, 2005). Therefore, patients will not be excluded from trial
participation based on the EGFR testing. The EGFR status will be assessed by the FISH
assay in biopsy or resection tissue samples and the test will be performed by a commercial
laboratory.