CHEMOTHERAPY/ALTERNATIVE
DELIVERY METHODS SECTION - LUNG CANCER TARGETED THERAPIES: COX-2
INHIBITORS, RETINOIDS, AND AEROSOLIZED DELIVERY
Ethan
Dmitrovsky, MD
Slide
1: Colleagues
DR.
DMITROVSKY: Thank you, Paul. I would, also, like to thank Diane
Bronzert and Janet Dancey for the invitation to speak today. In
addition I would like to thank these colleagues for contributing
to my presentation.
I
will be discussing the role of COX-2 inhibitors, retinoids and aerosolized
delivery in screening detected lung cancers. When our conference
began Jack Ruckdeschel spoke of the therapeutic nihilism that exists
in our field. The existence of novel pharmacologic agents that have
specific mechanisms of action represents an important resource to
address this nihilism. In fact, I think our challenge is actually
a little bit different. Today there are so many new compounds that
exist, we have to devise novel strategies to sort through these
compounds so that we can assess rapidly clinical activity. With
that as an introduction I would like to discuss these four objectives
today.
First,
as I said, I will speak about the role of COX-2 inhibitors, retinoids
and rexinoids which are RXR agonists in lung cancer therapy. I will
cite aerosolized delivery and make some additional points to those
that Jack had mentioned earlier, and I would like to highlight the
concept of mechanism-based combination therapy.
Finally, in the last portion of my brief presentation I will summarize
and propose new directions in this field. Here is the challenge.
This is a partial list of the new compounds that are available just
within the scope of the area that I will be discussing today, and
if we use all of these compounds it will be many years before we
determine whether any of these have activity. So, we must have alternative
strategies for using mechanistically-based therapeutic strategies
for lung cancers.
One
of the really attractive targets is cyclooxygenase 2. As we all
know, the advantage of selective COX-2 inhibition is the fact that
we can overcome the effects of constitutive COX-1 and get the preferential
effects of COX-2, the inducible form of cyclooxygenase 2. We have
several compounds that are now available for clinical use, and I
thought I would spend a few minutes describing why in lung cancer
in early stage lung cancer COX-2 inhibition may be an attractive
therapeutic target.
Here
are the points that I would like to summarize. I won't be showing
extensive primary data to validate these points, but they exist
in the literature. COX-2 is overexpressed in lung cancer. This is
a very common finding. COX-2 regulates synthesis of prostaglandins
that themselves can promote tumorigenesis. COX-2 inhibition reduces
carcinogen-mediated lung adenocarcinomas in the AJ mice, Jack spoke
about that model earlier. Also, COX-2 overexpression is well known
to inhibit apoptosis and to importantly regulate angiogenesis.
Here
is some primary data to show in Stage I lung cancers the differential
expression of COX-2 in tumors versus adjacent normal tissues. Here
are six cases. This is Western blot analysis, and you can see differential
expression of tumor versus adjacent normal tissues in several of
these cases.
This overexpression has been extended to a single cell level of
expression and I would like to actually show you the clinical correlation
for differential expression from Fadlo Khuri's work, a recent publication
from his group. You can see if you look at all stage I cases there
is a clear survival difference for stage I lung cancers for overexpression
of COX-2 cases versus those that have negative expression.
So,
to summarize the therapeutic role for COX-2 inhibition in early
stage lung cancer there is considerable preclinical and now clinical
findings that would suggest that COX-2 inhibition is quite appealing
for screening detected lung cancers. To optimize this therapeutic
approach the relevant mechanisms should be monitored in vivo and
again combination therapy with COX-2 inhibitors should clearly be
considered.
I
would like to move on to the next topic which is the role of retinoids
in lung cancer. As you well know, there is a long history for the
possible therapeutic role of retinoids in lung cancer therapy. The
first evidence suggested in 1925 by Walback and Howe that vitamin
A deficiency will cause squamous metaplasia in the lung quite reminiscent
of that that occurs in smokers. This is fully reversed by vitamin
A treatment. We know of the seminal work of Ki Hong and his colleagues
regarding 12-cis-retinoic acid in oral leukoplakia and Scott Lippman's
work showing that 13-cis plus interferon will treat several common
epithelial malignancies.
In
addition there were various Phase II trials that showed that certain
aerodigestive tract cancers, second cancers can be suppressed but
in smokers we now know both from the large Phase III trials with
beta-carotene and recent publication of Scott Lippman using 13-cis-retinoic
acid that in smokers there is no benefit and perhaps harm, while
there appears to be some benefit in those who are non-smokers. I
would argue that the mechanisms for these different biological effects
must be understood if we are going to advance the field in terms
of the role of retinoids in lung cancer therapy.
So,
we have been looking at these clinical trials from the perspective
of what we haven't learned. We don't yet know what the optimal dosage
is for using retinoids systemically. We don't know about the tissue
of pharmacokinetics. A point that has not been emphasized in these
trials to date is that we do not yet know whether RAR beta or other
markers are targeted, and we heard yesterday from the work of John
Minna and Adi Gazdar that in smokers people without cancer, RAR
beta suppression is extremely common. From the point of view of
retinoid signalling it is very important to emphasize this point
because these agents are transcriptional activators. They will not
work if the chromatin is not open. Methylation silences the chromatin,
prevents retinoic acid from acting, and we know that retinoic acid
treatment will not overcome this problem. So, the work of John Minna
and Adi Gazdar has shown that co-treatment with demethylation agents
plus retinoids can at least induce RAR beta expression. So, I think
from a retinoid perspective it is very important to emphasize the
point that chromatin must be open for these agents to act. How does
smoking interfere with these retinoid effects? All of these clinical
points underscore the need for mechanism driven lung cancer preventive
and therapeutic trials and the need to follow mechanisms in clinical
trials and choose agents based on their mechanism of action.
Slide
12: Derivation of Retinoid Resistant...Cells
In
the laboratory we have actually been studying how retinoic acid
and carcinogens might interact. We performed a very simple experiment
which was to take immortalized human bronchial epithelial cells
and try to develop retinoic acid resistant cells and this was a
very, very difficult thing to accomplish because retinoic acid resistance
was quite rare. How rare? Only 1 in 107 cells was retinoic acid
resistant. However, if we briefly treated bronchial epithelial cells
with carcinogen the level of resistance was dramatically increased
such that we were able to retrieve cells on the order of 1 in 1000
and 1 in 10,000 cells that were retinoic acid resistant. We would
argue that these cells are now very interesting resources to understand
how retinoids may work and how they may not work in patients who
are exposed to tobacco-specific carcinogens.
I
would like to underscore a couple of points that I have already
made and that is that the retinoic acid signaling pathway involves
two classes of receptors. The classical receptors are shown here,
the so-called "RAR" genes of which there are three, RAR
alpha, beta and gamma. So, RAR-beta is the receptor that has been
highlighted by the work of Rubin Lotan and Ki Hong as the transmitter
of the retinoic acid signal in diverse epithelial cells of the aerodigestive
tract.
This is the receptor that has been shown to be silenced through
methylation in a variety of epithelial tumors including lung cancers,
but all-trans-retinoic acid or 13-cis-retinoic acid only act through
the classical pathway. They do not affect the non-classical retinoids,
the so-called "rexinoid" pathway RXR alpha, beta and gamma.
To activate this pathway you either need a bifunctional compound
like 9-cis-retinoic acid that activates both pathways or alternatively
you need a specific agonist for the RXR pathway. Now, why would
a rexinoid and RXR agonist be active in lung carcinogenesis? One
reason is it bypasses the defect of RAR beta suppression. Another
reason that this may be active is that it itself may signal a distinct
set of activated or repressed genes that themselves would have therapeutic
activity.
We have some clinical evidence that this hypothesis is valid and
that is from the work of Fadlo Khuri who recently published results
of survival in patients who had been treated with a rexinoid, bexarotene,
plus chemotherapy and surprisingly a very prominent tail of survival
is noted in roughly 30% of the patients. So, here we have in vivo
clinical evidence that the retinoids might have activity if we use
the RXR agonists.
Another
role for the retinoids in therapy of these early lesions would be
in the setting of aerosolized delivery and Jack Roth has mentioned
the studies that have been led by Jim Mulshine of carcinogen-induced
tumors in the HA mouse model where aerosolized retinoids were able
to suppress dramatically the incidence of these pulmonary adenomas.
Although some toxicity was shown at very, very high dosages it is
important that this targeted approach be considered for the use
of small molecules, agonists such as the retinoids themselves.
This is the same slide that Jack showed you before, emphasizing
the value of targeted therapy, the efficiency of delivery and the
reduction of systemic toxicity.
I would like to emphasize two points in addition to those that Jack
had mentioned. First of all we don't yet know what agent to put
into this reservoir. Second point is that the aerosolized delivery
turns out to be a fairly complex way of delivering therapeutic agents.
Why? Because the size of the particles really will determine where
in the lung these agents will be targeted. So, if we want to target
central versus peripheral lesions we will have to program the size
of the aerosolized particles. It is a point that really requires
extensive preclinical study and even clinical study before we can
have the hope of having reservoirs that people will take with them
to suppress carcinogenesis in the lung.
We
have to broaden our scope of the potential pharmacologic agents
that can be used and as Jack had mentioned the work of Jonathan
Curry using Tyler Jacks' ras knock-in mouse we can see that vectors
might preferentially be taken up by tumors in the lung shown histologically
here, and there is really tremendous enthusiasm for aerosolized
delivery but there is much more that we need to learn on a preclinical
basis.
I
would like to now emphasize the next concept that I wanted to discuss
today which is combination therapy and I am borrowing from some
preclinical studies of Mark Erasi and others at Colorado to make
this point.
The
point to this slide is to make the concept that combination therapy
with chemotherapy agents using a novel agent like exisulind might
lead to cooperative effects shown here in terms of the marked suppression
of tumorigenesis in the lung with a major survival advantage that
I won't show you of these rats, but that each of these agents act
through different pathways.
The
fact that they act through different pathways is underscored in
the next two slides which show the pro-apoptotic effects of exisulind,
the growth suppressor effects of taxotere leading to cooperativity
in terms of the therapeutic outcome, and this cooperativity is shown,
also, in the antiproliferative effects when we consider the effects
of growth suppression of these different agents.
So,
how do we move forward in this field, and that is what I would like
to discuss in the next few minutes. I would like to emphasize two
points, and that is first the value of preclinical models to highlight
therapeutic pathways, secondly that there is a need to validate
whether pathways that are targeted in vitro are also targeted in
vivo in clinical subjects that are entered onto hypothesis driven
trials.
We
have spent considerable time trying to understand in the lung which
pathways are critical for the maintenance or progression of a bronchial
epithelial cell transforming into a fully malignant phenotype. Over
several years we have actually exploited a very, very simple model
that I would like to discuss with you. Curt Harris was kind enough
to provide our laboratory his immortalized human bronchial epithelial
cell line called BEAS-2B and Jonathan Langenfeld and others in the
laboratory exposed these cells for a very brief period of time,
just for a day to the carcinogen NNK and over the course of several
months we were able to generate a lung cancer. However, if we co-treated
with retinoic acid we could block this transformation and essentially
develop a chemopreventive phenotype. Over several years we have
published a mechanism that is associated with this prevention and
I would like to just summarize this mechanism.
We
found a major difference between the transformed and the chemoprevented
cells was the effects exerted upon the cell cycle at a very precise
point and this is at the G1-S phase transition.
The pathway that we discovered was unexpected to us a post-transcriptional
mechanism, a translational effect. That is the retinoids prevent
transformation of human bronchial epithelial cells at least partly
by causing G1 arrest through a proteolysis of the cyclins, cyclin
D1 and cyclin E. This permits repair of genomic damage because there
is delay of the transition from G1 through S much like what P53
does. So, we have found that retinoids will do this, and we are
now completing a very detailed structure function analysis to see
whether the retinoids are the optimal agents that affect this pathway.
We decided to come at the model from a different perspective. Valerie
Rusch and I had published several years ago that the epidermal growth
factor receptor was frequently overexpressed in preneoplastic lesions
of the lung. So, we asked would this model select for an EGFR- dependent
pathway and would chemopreventive cells repress this activation?
So,
Fulvio Lenardo has studied this in the laboratory and quite interestingly
we found that carcinogens themselves will activate not only EGFR
pathway but also coincidentally to the transformation phenotype.
However, when we chemoprevent these cells with the agent retinoic
acid EGF receptor is dramatically repressed.
One
slide that shows you the primary data of this. Here are the transformed
cells and here are the chemoprevented cells. This is a phospho-specific
tyrosine recognizing antibody before and after EGF treatment. You
can see the dramatic stimulation of autophosphorylation of the epidermal
growth factor receptor in the transformed cells treated with EGF
but not the chemopreventive and this is due to a major suppression
of levels of epidermal growth factor receptor.
So, here we have a pathway that has been validated in vitro, that
also appears to be active in vivo and I would argue it is an attractive
pathway to target for early lung carcinogenesis.
This
simple model we have, also, used in another way, and I would like
to describe that. So, here we have unique reagents to examine differences
between transformed and chemoprevented cells and I have joined forces
with Marian Mysic at Dartmouth. She is a physicist who has been
studied laser imaging and auto-fluorescence, and we are all familiar
pioneering work of Steve Lamb in this area. We have developed an
alternative strategy for discriminating between transformed and
chemoprevented as well as immortalized cells, and this technology
we have termed fluorescence lifetime spectrometry.
This technology would allow us to identify central preneoplastic
lesions and we have heard a lot about peripheral lesions, and I
would like to just make a comparison to the so-called "life
bronchoscopy" that Steve Lamb has led the way on, and here
are the key differences. Lifetime spectroscopy allows only spectral
resolution. The FLS technology that we are working with allows spectral
and temporal resolution, and you can see that the excitation with
the FLS technology is over a very wide range of wavelength.
So, how is this approach used and we have been able to identify
in a recent publication from our laboratory, the endogenous chromophors
that actually lead to the auto-fluorescence and what we find here
and here is a pseudo image of confocal microscopy of the endogenous
NADH and co-localization of the endogenous flavins that are the
cause for the auto-fluorescence. In our recent publication we have
shown that we can discriminate between immortalized and transformed
cells on the basis of this auto-fluorescence pattern. We have recently
received funding to build the first prototype device and we have
actually entered the first patients into bronchoscopy directed imaging
studies at Dartmouth and hopefully in the near future we will be
able to discuss those results.
So
what have I said to you today? I have discussed that there exists
a strong rationale for the use of selective COX-2 inhibitors, especially
non-classical retinoids in screening detected lung cancers. There
is a role for combination regimens that should investigate agents
that activate different pathways and finally I would strongly argue
for the need of mechanisms that should be studied in proof of principle
clinical trials so that we can validate whether pathways identified
in vitro are, also, activated in vivo.
So,
I would like to end my presentation by returning to its beginning,
and remind you that we have a surplus of compounds to study, and
our challenge is really quite different. The challenge that we have
at hand is to establish a way for sorting out the many compounds
that are potentially clinically active and then validate their activity
in mechanism-driven clinical trials, and that is really a challenge
for all of us in this room to identify how these agents act and
when they should be applied in the treatment of early stage lung
cancer. Thanks for your attention.
DR.
DMITROVSKY: Thank you, Paul. I would, also, like to thank Diane
Bronzert and Janet Dancey for the invitation to speak today. In
addition I would like to thank these colleagues for contributing
to my presentation.
I
will be discussing the role of COX-2 inhibitors, retinoids and aerosolized
delivery in screening detected lung cancers. When our conference
began Jack Ruckdeschel spoke of the therapeutic nihilism that exists
in our field. The existence of novel pharmacologic agents that have
specific mechanisms of action represents an important resource to
address this nihilism. In fact, I think our challenge is actually
a little bit different. Today there are so many new compounds that
exist, we have to devise novel strategies to sort through these
compounds so that we can assess rapidly clinical activity. With
that as an introduction I would like to discuss these four objectives
today.
First,
as I said, I will speak about the role of COX-2 inhibitors, retinoids
and rexinoids which are RXR agonists in lung cancer therapy. I will
cite aerosolized delivery and make some additional points to those
that Jack had mentioned earlier, and I would like to highlight the
concept of mechanism-based combination therapy.
One
of the really attractive targets is cyclooxygenase 2. As we all
know, the advantage of selective COX-2 inhibition is the fact that
we can overcome the effects of constitutive COX-1 and get the preferential
effects of COX-2, the inducible form of cyclooxygenase 2. We have
several compounds that are now available for clinical use, and I
thought I would spend a few minutes describing why in lung cancer
in early stage lung cancer COX-2 inhibition may be an attractive
therapeutic target. 
Here
are the points that I would like to summarize. I won't be showing
extensive primary data to validate these points, but they exist
in the literature. COX-2 is overexpressed in lung cancer. This is
a very common finding. COX-2 regulates synthesis of prostaglandins
that themselves can promote tumorigenesis. COX-2 inhibition reduces
carcinogen-mediated lung adenocarcinomas in the AJ mice, Jack spoke
about that model earlier. Also, COX-2 overexpression is well known
to inhibit apoptosis and to importantly regulate angiogenesis. 
Here
is some primary data to show in Stage I lung cancers the differential
expression of COX-2 in tumors versus adjacent normal tissues. Here
are six cases. This is Western blot analysis, and you can see differential
expression of tumor versus adjacent normal tissues in several of
these cases. 
This overexpression has been extended to a single cell level of
expression and I would like to actually show you the clinical correlation
for differential expression from Fadlo Khuri's work, a recent publication
from his group. You can see if you look at all stage I cases there
is a clear survival difference for stage I lung cancers for overexpression
of COX-2 cases versus those that have negative expression. 
So,
to summarize the therapeutic role for COX-2 inhibition in early
stage lung cancer there is considerable preclinical and now clinical
findings that would suggest that COX-2 inhibition is quite appealing
for screening detected lung cancers. To optimize this therapeutic
approach the relevant mechanisms should be monitored in vivo and
again combination therapy with COX-2 inhibitors should clearly be
considered.
I
would like to move on to the next topic which is the role of retinoids
in lung cancer. As you well know, there is a long history for the
possible therapeutic role of retinoids in lung cancer therapy. The
first evidence suggested in 1925 by Walback and Howe that vitamin
A deficiency will cause squamous metaplasia in the lung quite reminiscent
of that that occurs in smokers. This is fully reversed by vitamin
A treatment. We know of the seminal work of Ki Hong and his colleagues
regarding 12-cis-retinoic acid in oral leukoplakia and Scott Lippman's
work showing that 13-cis plus interferon will treat several common
epithelial malignancies. 
In
addition there were various Phase II trials that showed that certain
aerodigestive tract cancers, second cancers can be suppressed but
in smokers we now know both from the large Phase III trials with
beta-carotene and recent publication of Scott Lippman using 13-cis-retinoic
acid that in smokers there is no benefit and perhaps harm, while
there appears to be some benefit in those who are non-smokers. I
would argue that the mechanisms for these different biological effects
must be understood if we are going to advance the field in terms
of the role of retinoids in lung cancer therapy. 
So,
we have been looking at these clinical trials from the perspective
of what we haven't learned. We don't yet know what the optimal dosage
is for using retinoids systemically. We don't know about the tissue
of pharmacokinetics. A point that has not been emphasized in these
trials to date is that we do not yet know whether RAR beta or other
markers are targeted, and we heard yesterday from the work of John
Minna and Adi Gazdar that in smokers people without cancer, RAR
beta suppression is extremely common. From the point of view of
retinoid signalling it is very important to emphasize this point
because these agents are transcriptional activators. They will not
work if the chromatin is not open. Methylation silences the chromatin,
prevents retinoic acid from acting, and we know that retinoic acid
treatment will not overcome this problem. So, the work of John Minna
and Adi Gazdar has shown that co-treatment with demethylation agents
plus retinoids can at least induce RAR beta expression. So, I think
from a retinoid perspective it is very important to emphasize the
point that chromatin must be open for these agents to act. How does
smoking interfere with these retinoid effects? All of these clinical
points underscore the need for mechanism driven lung cancer preventive
and therapeutic trials and the need to follow mechanisms in clinical
trials and choose agents based on their mechanism of action. 
In
the laboratory we have actually been studying how retinoic acid
and carcinogens might interact. We performed a very simple experiment
which was to take immortalized human bronchial epithelial cells
and try to develop retinoic acid resistant cells and this was a
very, very difficult thing to accomplish because retinoic acid resistance
was quite rare. How rare? Only 1 in 107 cells was retinoic acid
resistant. However, if we briefly treated bronchial epithelial cells
with carcinogen the level of resistance was dramatically increased
such that we were able to retrieve cells on the order of 1 in 1000
and 1 in 10,000 cells that were retinoic acid resistant. We would
argue that these cells are now very interesting resources to understand
how retinoids may work and how they may not work in patients who
are exposed to tobacco-specific carcinogens. 
I
would like to underscore a couple of points that I have already
made and that is that the retinoic acid signaling pathway involves
two classes of receptors. The classical receptors are shown here,
the so-called "RAR" genes of which there are three, RAR
alpha, beta and gamma. So, RAR-beta is the receptor that has been
highlighted by the work of Rubin Lotan and Ki Hong as the transmitter
of the retinoic acid signal in diverse epithelial cells of the aerodigestive
tract.
We have some clinical evidence that this hypothesis is valid and
that is from the work of Fadlo Khuri who recently published results
of survival in patients who had been treated with a rexinoid, bexarotene,
plus chemotherapy and surprisingly a very prominent tail of survival
is noted in roughly 30% of the patients. So, here we have in vivo
clinical evidence that the retinoids might have activity if we use
the RXR agonists.
Another
role for the retinoids in therapy of these early lesions would be
in the setting of aerosolized delivery and Jack Roth has mentioned
the studies that have been led by Jim Mulshine of carcinogen-induced
tumors in the HA mouse model where aerosolized retinoids were able
to suppress dramatically the incidence of these pulmonary adenomas.
Although some toxicity was shown at very, very high dosages it is
important that this targeted approach be considered for the use
of small molecules, agonists such as the retinoids themselves. 
This is the same slide that Jack showed you before, emphasizing
the value of targeted therapy, the efficiency of delivery and the
reduction of systemic toxicity.
We
have to broaden our scope of the potential pharmacologic agents
that can be used and as Jack had mentioned the work of Jonathan
Curry using Tyler Jacks' ras knock-in mouse we can see that vectors
might preferentially be taken up by tumors in the lung shown histologically
here, and there is really tremendous enthusiasm for aerosolized
delivery but there is much more that we need to learn on a preclinical
basis. 
I
would like to now emphasize the next concept that I wanted to discuss
today which is combination therapy and I am borrowing from some
preclinical studies of Mark Erasi and others at Colorado to make
this point. 
The
point to this slide is to make the concept that combination therapy
with chemotherapy agents using a novel agent like exisulind might
lead to cooperative effects shown here in terms of the marked suppression
of tumorigenesis in the lung with a major survival advantage that
I won't show you of these rats, but that each of these agents act
through different pathways.
They
have been using a nude mouse, nude rat model for looking at A549
adenocarcinomas that will readily metastasize to the lung. 
The
fact that they act through different pathways is underscored in
the next two slides which show the pro-apoptotic effects of exisulind,
the growth suppressor effects of taxotere leading to cooperativity
in terms of the therapeutic outcome, and this cooperativity is shown,
also, in the antiproliferative effects when we consider the effects
of growth suppression of these different agents. 
Here
you can see clear cooperativity when the two agents are combined.
So,
how do we move forward in this field, and that is what I would like
to discuss in the next few minutes. I would like to emphasize two
points, and that is first the value of preclinical models to highlight
therapeutic pathways, secondly that there is a need to validate
whether pathways that are targeted in vitro are also targeted in
vivo in clinical subjects that are entered onto hypothesis driven
trials.
We
have spent considerable time trying to understand in the lung which
pathways are critical for the maintenance or progression of a bronchial
epithelial cell transforming into a fully malignant phenotype. Over
several years we have actually exploited a very, very simple model
that I would like to discuss with you. Curt Harris was kind enough
to provide our laboratory his immortalized human bronchial epithelial
cell line called BEAS-2B and Jonathan Langenfeld and others in the
laboratory exposed these cells for a very brief period of time,
just for a day to the carcinogen NNK and over the course of several
months we were able to generate a lung cancer. However, if we co-treated
with retinoic acid we could block this transformation and essentially
develop a chemopreventive phenotype. Over several years we have
published a mechanism that is associated with this prevention and
I would like to just summarize this mechanism.
We
found a major difference between the transformed and the chemoprevented
cells was the effects exerted upon the cell cycle at a very precise
point and this is at the G1-S phase transition.
So,
Fulvio Lenardo has studied this in the laboratory and quite interestingly
we found that carcinogens themselves will activate not only EGFR
pathway but also coincidentally to the transformation phenotype.
However, when we chemoprevent these cells with the agent retinoic
acid EGF receptor is dramatically repressed. 
One
slide that shows you the primary data of this. Here are the transformed
cells and here are the chemoprevented cells. This is a phospho-specific
tyrosine recognizing antibody before and after EGF treatment. You
can see the dramatic stimulation of autophosphorylation of the epidermal
growth factor receptor in the transformed cells treated with EGF
but not the chemopreventive and this is due to a major suppression
of levels of epidermal growth factor receptor.
This
simple model we have, also, used in another way, and I would like
to describe that. So, here we have unique reagents to examine differences
between transformed and chemoprevented cells and I have joined forces
with Marian Mysic at Dartmouth. She is a physicist who has been
studied laser imaging and auto-fluorescence, and we are all familiar
pioneering work of Steve Lamb in this area. We have developed an
alternative strategy for discriminating between transformed and
chemoprevented as well as immortalized cells, and this technology
we have termed fluorescence lifetime spectrometry.
This technology would allow us to identify central preneoplastic
lesions and we have heard a lot about peripheral lesions, and I
would like to just make a comparison to the so-called "life
bronchoscopy" that Steve Lamb has led the way on, and here
are the key differences. Lifetime spectroscopy allows only spectral
resolution. The FLS technology that we are working with allows spectral
and temporal resolution, and you can see that the excitation with
the FLS technology is over a very wide range of wavelength. 
So, how is this approach used and we have been able to identify
in a recent publication from our laboratory, the endogenous chromophors
that actually lead to the auto-fluorescence and what we find here
and here is a pseudo image of confocal microscopy of the endogenous
NADH and co-localization of the endogenous flavins that are the
cause for the auto-fluorescence. In our recent publication we have
shown that we can discriminate between immortalized and transformed
cells on the basis of this auto-fluorescence pattern. We have recently
received funding to build the first prototype device and we have
actually entered the first patients into bronchoscopy directed imaging
studies at Dartmouth and hopefully in the near future we will be
able to discuss those results. 
So
what have I said to you today? I have discussed that there exists
a strong rationale for the use of selective COX-2 inhibitors, especially
non-classical retinoids in screening detected lung cancers. There
is a role for combination regimens that should investigate agents
that activate different pathways and finally I would strongly argue
for the need of mechanisms that should be studied in proof of principle
clinical trials so that we can validate whether pathways identified
in vitro are, also, activated in vivo. 
So,
I would like to end my presentation by returning to its beginning,
and remind you that we have a surplus of compounds to study, and
our challenge is really quite different. The challenge that we have
at hand is to establish a way for sorting out the many compounds
that are potentially clinically active and then validate their activity
in mechanism-driven clinical trials, and that is really a challenge
for all of us in this room to identify how these agents act and
when they should be applied in the treatment of early stage lung
cancer. Thanks for your attention.