FUTURE STRATEGIES IN LUNG CANCER PREVENTION

After more than 2 decades of unsuccessful interventions in
NSCLC, evidence has now accumulated to suggest that approaches
relying exclusively on the development of epidemiologic
guidance for selection of compounds are unlikely to lead
to effective interventions across broad and disparate populations.
Although clues continue to emerge from the epidemiologic
literature, including a study by Schabath et al. 74 showing
that dietary phytoestrogens appear to significantly reduce lung
cancer risk, few have suggested moving aggressive prevention
approaches forward without further, well-designed, biomarkerdriven
trials.
Current targeted approaches focus on several pathways. To
date, significant evidence exists to show that overexpression of
the EGFR occurs throughout lung carcinogenesis, 16,75,76 and
that mutations of the EGFR tyrosine kinase (EGFR-TK) binding
domain can be seen diffusely across the normal-appearing
airway in patients with resected adenocarcinomas who are nonsmokers
and harbor an EGFR-TK binding domain mutation in
their primary tumor. This discovery by Tang et al. 77 of a novel
field effect in which EGFR-TK mutations are seen in apparently
normal-appearing, nontobacco damaged airways suggests
that despite the absence of an identifiable carcinogen leading to
EGFR mutations, this represents a new and as yet etiologically
unclear type of field effect. Several investigators have proposed
novel chemoprevention approaches using the EGFR-TK inhibitors,
erlotinib, and gefitinib, in high-risk patient populations.
Given the fact that the incidence of EGFR-TK mutation- bearing
adenocarcinoma of the lung is most frequently seen in Asian,
female nonsmokers with lung cancer, 78,79 the identification of
those individuals at greatest risk would likely lead to recognition
of a target population, thereby enhancing the likelihood of a successful
intervention with an EGFR-TKI.
Important data also implicate the progressive upregulation
of cyclooxygenase-2 (COX-2) in lung carcinogenesis, and exciting
preliminary trials targeting both suppression of COX-2 with
selective COX-2 inhibitors by Mao et al. 80 and Kim et al. 81 have
shown provocative results to date. Another approach taken by the
University of Colorado group has sought to upregulate prostacycline,
thereby downregulating prostaglandin E2 (PGE-2), potentially
the critical downstream effector pathway for COX-2. 82
Another promising targeted approach is based on epidemiologic
data. Govindarajan et al. 83 showed that thiazolidinediones (TZD),
peroxisome proliferator-activated receptor (PPAR) stimulators,
are able to induce cell cycle arrest. These investigators examined a
large cohort study in 87,678 male veterans older than 40 years, and
showed a 33% reduction of lung cancer risk among the 11,289
TZD users. Although the study failed to account for variations by
smoking status, this represents a potentially exciting and useful
chemopreventive approach. As reviewed by Nemenoff, 84,85 activation
of PPAR inhibits lung tumorigenesis as demonstrated by
animal studies in which increased PPAR may be chemopreventive
against developing lung tumors. In established lung canc er,
PPAR- activation inhibits proliferation, induces apoptosis, and
promotes a less invasive phenotype by promoting epithelial differentiation,
and perhaps blocking epithelial mesenchymal transition.
PPAR- inhibition of chemokine production may also negatively
impact tumor progression and metastasis. Although activation of
PPAR- can occur by direct binding of pharmacological ligands
to the molecule, emerging data indicate that PPAR - activation
can occur through engagement of other signal transduction pathways,
including Wnt signaling and prostaglandin production.
Defining the molecular targets of TZDs mediating a specific response
will be critical in the further development of second-generation
PPAR - drugs. Cardiac events have been recorded with
the use of TZDs, and the development of more selective PPAR -
activators could potentially be therapeutically effective, without
leading to adverse cardiac events.
Other methods have involved utilizing corticosteroids,
both inhaled and oral, or inhaled retinoids, all of which have
shown promise. Unfortunately, two randomized phase II studies
in high-risk patients have failed to show a trend in favor of
inhalational corticosteroids. 86,87
While these and other data continue to be attractive, a
surprising trial that sought to prevent SPT skin cancer in a
selenium-poor population in Arizona resulted in the fortuitous
finding of a dramatic reduction in lung cancer incidence
(34% reduction). 38 The phase III Eastern Cooperative
Oncology Group (ECOG) selenium trial, E5597, has accrued
approximately 1300 of 1900 planned patients. This
trial randomizes patients in a 2:1 ratio to receive selenium
methionine at 200 g/day versus placebo. Patients must have
the identical criteria to the prior intergroup study, that is, a
history of stage I to II NSCLC fully resected, patients being
registered between 6 weeks and up to 3 years after surgery.
Whether the selenium trial ultimately yields definitive answers,
the recommendation must be to continue to build biomarker-
driven, targeted, phase II approaches that appropriately
pursue modulation of a critical biomarker, and confirm
it prior to proceeding with aggressive, broad, studies in large
patient populations. Once the appropriate biomarkers have
been shown to be consistently modulated in selected populations,
only then can we proceed with confidence to develop
targeted agents that can meaningfully reduce the incidence of
lung cancer in at-risk populations.