EXHIBIT 99.2
On February 8, 2005, OSI Pharmaceuticals, Inc. (the "Company") held a webcast
conference call regarding its financial results for the transition period ended
December 31, 2004 as well as an update on the Company's business. The following
represents a textual representation of certain portions of the transcript of the
webcast conference call consisting of remarks by Colin Goddard, Chief Executive
Officer of the Company, and Robert Van Nostrand, Vice President and Chief
Financial Officer of the Company.
OSI PHARMACEUTICALS
MODERATOR: COLIN GODDARD
FEBRUARY 8, 2005
7:00 A.M. CT
[Operator's Introduction]
[Colin Goddard's Introduction]
[Robert Van Nostrand] Thank you Colin. Good morning everyone. Before I start
with the review of our financial results for the transition period ended
December 31st, 2004, I'd like to remind you that we will be making
forward-looking statements relating to financial results and clinical and
regulatory developments on the call today. These statements are subject to
various risks and uncertainties that could cause the results to differ
materially from those expressed in any forward-looking statements. I'll
refer you to our SEC disclosure documents for a detailed description of
the risk factors affecting our business.
[Robert Van Nostrand discusses quarterly information.]
We anticipate that our total corporate revenues for 2005 will be between a
conservative $105 and $170 million. This large range reflects the impact
of Tarceva on our P&L statement, and the difficulty in providing guidance
at this early stage of the Tarceva launch. We expect to fine-tune our
guidance as the year progresses. We have agreed
with our partners not to provide individual product revenue or brand sales
and marketing guidance for Tarceva, and not to comment on pricing
strategy, upon which Genentech has the final decision following input from
our commercial team.
On the expense side for the upcoming year, we anticipate research and
development expense to be in the range of $120 to $130 million, and our
selling, general, and administrative expense to be between $90 and $100
million. We continue to work on a goal of containing our overall core
operating costs, essentially R&D plus SG&A close to $225 million, until we
clearly see the trajectory of Tarceva sales. We anticipate our 2005 loss
per share will improve to between $1.05 and $2.35, versus the $6.50 loss
per share reported for fiscal 2004.
Finally, having updated our internal mid-course Tarceva revenue
projections, we are moving on goal for quarter-by-quarter profitability,
up from the original goal of within three years of launch, to a new goal
of attaining quarter-by-quarter profitability by the end of 2006. It
should be noted, however, that our guidance on operating expense does not
include any effects of implementing FASB 123R regarding the expensing of
stock options, which will be effective for periods after June 15, 2005. We
are currently reviewing all alternatives for calculating the expense,
although we do not expect this to significantly impact our timeline to
profitability.
[Colin Goddard] Thanks Bob. We recognize the Tarceva launch is of paramount
importance to our investors, and although we've agreed with our partners
to limit the amount of brand-
specific guidance that we provide, we do believe it's important and
helpful for us to provide some thoughts on how we view the emerging EGFR
market, and also to update investors on the progress of the Tarceva
program, and events that could impact the extent to which our corporate
revenues will reach the upper end of the broad range given.
As you're aware, the FDA approved Tarceva on November the 18th of last
year for the treatment of advanced non-small-cell lung cancer patients
after failure of at least one prior chemotherapy regimen. The approval
reflects the robust survival benefit demonstrated in the BR21 study, and
the label contains no restrictions on the use of Tarceva in the second-
and third-line setting. This contrasts with the third-line accelerated
approval label granted for gefitinib, the principle competitor to Tarceva
in this market, which was based on response rate data alone.
Our competitor announced in December that gefitinib failed to demonstrate
an overall survival advantage in a large randomized Phase III trial, the
ISEL study, that was similar in design to our BR21 trial. Nonetheless, the
market performance of gefitinib during 2004 provides a useful benchmark
from which to consider the Tarceva launch. By the end of the third quarter
of 2004, prior to the Tarceva approval and the announcement of the ISEL
failure, sales of gefitinib was an annualized run rate of approximately
$240 million per year in the U.S. We believe that this reflects good
penetration into the third-line non-small-cell lung setting, but less
penetration for off-label use in earlier stage disease.
We believe that the full approval for the use of Tarceva in the second-
and third-line setting, and the failure of gefitinib to demonstrate a
statistically significant survival benefit in the analogous ISEL study
positions us well to take existing market share from our competitor. In
fact, survey data suggests that even before the ISEL news broke, we've
captured an approximately 40 percent share of new patients scripts, and as
of January 28th, just 10 weeks post-launch, this share had increased to
over 80 percent. We would anticipate this dominance carrying through to
total scripts over the next several quarters.
Clearly this will be affected somewhat by the ultimate fate of gefitinib
following the upcoming ODAC meeting, and FDA's full review of the ISEL
data, the extent to which the observed switching from gefitinib to Tarceva
continues, and our ability to clearly differentiate Tarceva in the
marketplace.
There's also some evidence that the total number of EGFR scripts is
rising. Survey data shows approximately 2,600 weekly scripts for the EGFR
class in the period following Tarceva launch, excluding the holiday weeks,
versus approximately 2,200 weekly scripts in the period prior to the
Tarceva launch. We believe that this is a credible performance, given some
confusion in the marketplace following the ISEL data.
We believe that our success in growing total EGFR scripts in this launch
year will largely be dependent upon the following factors -- our ability
to continue to educate the oncology community to the differences between
gefitinib and Tarceva, thereby
overcoming any perceived class taint of the Tarceva data with the recently
reported gefitinib data; our ability to address what we believe to be are
misconceptions within some segments of the oncology community concerning
the broad potential of the drug in treating lung cancer patients; our
ability to convince oncologists of the value of using Tarceva in the
second-line setting in place of chemotherapy, early use in pancreatic
cancer, and the speed of the pancreatic cancer regulatory process; and the
impact of emerging data presented at upcoming major cancer meetings to
further broaden Tarceva use.
Based largely on data from studies with gefitinib, some researchers have
advanced the notion that the EGFR class of agents are only effective in
small subsets of lung cancer patients who are of Asian origin, who have
tumors with mutations in the EGFR gene, or are patients who have never
smoked. We have consistently argued that this significantly understates
the value of Tarceva treatments in lung cancer patients, while we continue
to believe that the ability to safely dose the drug at exposure levels
appreciably higher than the competitor's agents results in a demonstrated
survival benefit for most non-small-cell lung cancer patients, including
those whose tumors involve non-mutated, or wild-type EGFR, and patients
who are current or former smokers. We continue to study these aspects of
Tarceva use in non-small-cell lung cancer patients.
Exploratory univariate subset analysis of the data from the BR21 study had
indicated a trend toward a statistically significant survival benefit in
current or former smokers. Note that the BR21 study was not powered to
show statistical significance in any of the
subsets. Smoking was not a stratification factor for randomization in the
BR21 study, and this led to imbalances in patient subsets with respect to
smoking status. We have, therefore, conducted a series of further
statistical analysis using multi-barrier COX models. This approach reveals
a statistically significant hazard ratio of .81, or a 23 percent
improvement in survival for current or former smokers.
Pharmacokinetic studies of samples from the BR21 study also suggested that
smokers were clearing Tarceva quicker, and therefore receiving less
exposure to drug. Given our belief that dose and exposure are important
components of effective Tarceva therapy, we conducted a clinical trial in
healthy volunteer patients designed to assess the impact of smoking on
blood levels of Tarceva. Preliminary results from this study, which will
be presented in detail at the upcoming AACR meeting to be held in Anaheim,
California during April 2005, have indicated that the levels of drug in
the blood of smokers following a single dose of Tarceva are indeed
appreciably lower than those achieved in nonsmokers.
Studies are also nearing completion on the EGFR mutation status, EGFR
wild-type status, and overall EGFR suppression status of tumors specimens
obtained in both the BR21 non-small-cell lung cancer Phase III trial, and
the pancreatic cancer PA3 trial. We anticipate the results from all of
these studies will be presented at the upcoming ASCO meeting in Orlando
during May 2005, and that they will shed further light on the question of
whether it is possible, or even appropriate, to select subsets of lung
cancer patients for therapy with Tarceva using the available unvalidated
methodologies.
In this respect we have always believed that the majority of
non-small-cell lung cancer patients presenting for treatment with Tarceva
will have no available sample for EGFR testing, and that
immunohistochemistry, or IHC, is not an ideal, nor validated, means by
which to select lung cancer patients for therapy with Tarceva. Our label
requires no testing prior to the initiation of therapy with Tarceva.
Nonetheless, there is continued interest in IHC testing, not only in lung
cancer, but also in colon cancer where the wisdom of using IHC testing for
patient selection with another EGFR targeted agent was questioned in a
recently published article in the "Journal of Clinical Oncology."
As a result, we have continued to study results and data from the BR21
trial on patients determined to be EGFR positive or negative by IHC
assessment of tumor tissue samples. After extraordinary efforts to recover
and collect archived tumor samples, we have now been able to assess EGFR
status on 45 percent of the participants in the BR21 study versus the 33
percent of participants reported in the label. Depending on the criteria
applied between 56 and 70 percent of these patients were EGFR positive,
and using the criteria described in the label, the hazard ratio for
survival in the EGFR negative group falls to 0.93, compared to a hazard
ratio of 1.01 in the label, although neither number is statistically
significant.
This, coupled to a response rate of 3.8 percent for this EGFR negative
group, supports our contention that we should not rule out a modest
patient benefit in this group. We continue to believe that IHC testing is
an imperfect tool for patient selection. Further,
given the relatively mild side effect profile elicited by Tarceva, the
fact that the large EGFR unknown subset in the BR21 study derived a
comparable survival benefit to the overall population, the time and cost
involved in either recovering archived samples, or obtaining fresh biopsy
material, and the fact that we do not believe it is possible to rule out a
modest patient benefit in the EGFR negative subgroup, we do not believe
that IHC testing will have a significant impact on the use of Tarceva in
lung cancer patients.
The main near-term opportunity for us to expand the total EGFR market is
in the second-line setting, not only by displacing chemotherapy, but by
also expanding the addressable pool of second-line patients. Survey data
has indicated that there are between 41,000 and 43,000 non-small-cell lung
cancer patients who undergo second- and third-line therapy with cytotoxic
chemotherapy. We believe this number might understate the number of
patients available for treatment after the failure of front-line
chemotherapy, because physicians and patients who are unwilling to
undertake the further rigors of continued conventional chemotherapy may
consider therapy with Tarceva, an oral once-daily treatment with a
relatively mild side effect profile. We've always anticipated that
displacing chemotherapy use in the second-line setting would save time,
but continue to believe that the combination of comparable efficacy and
better side effect profile will favor expanded Tarceva use.
Clearly the fact that Tarceva has also demonstrated a survival benefit in
a large Phase III trial in pancreatic cancer is both a major
differentiating factor for the drug, and noteworthy in that Tarceva is the
first non-chemotherapy agent and only the second drug
ever to demonstrate a survival benefit, which was seen in combination with
gemcitabine in this setting. The data from the PA3 study, showing a 23.5
percent improvement in overall survival for patients receiving Tarceva
plus gemcitabine, versus those receiving gemcitabine plus placebo, was
seen in front-line patients with locally advanced or metastatic pancreatic
cancer. The data was recently presented at the ASCO GI meeting in Florida.
The data was well received, and the supplemental NDA filing is clearly a
high priority for the organization. We anticipate completing the filing in
the May timeframe, and have already been notified by the FDA that the
single study will be acceptable for submission. We would anticipate
receiving rapid review of this application and estimate approval by
year-end, assuming a smooth and successful review with the FDA. With
limited available therapeutic choices in pancreatic cancer, we expect to
achieve good market uptake in this setting. There are approximately 32,000
newly diagnosed cases of pancreatic cancer in the U.S. each year.
In addition to the EGFR mutation and expression data from the BR21 and PA3
studies, we also expect to see additional data at this year's ASCO
meeting, including updated data from ongoing trials combining Tarceva and
Avastin, and data from monotherapy Tarceva trials in the front-line
non-small-cell lung cancer setting. On balance, we consider the
preponderance of factors to be supportive of a solid Tarceva launch year,
with more reasons to believe that Tarceva revenues will tend towards the
upper end of our range than not.
Even though launch year sales may be impacted by some currently unknown
factors, such as the regulatory fate of gefitinib in non-small-cell lung
cancer, the extent of our ability to replace second-line chemotherapy in
the near-term, and the speed of approval in pancreatic cancer, we continue
to believe that Tarceva will emerge as the dominant EGFR inhibitor in
non-small-cell lung cancer, that the emerging science will support our
view on the full potential of the drug, and that Tarceva will enjoy a
successful launch in pancreatic cancer.
When we also consider the impact of Tarceva approval in other territories,
pricing strategy, the introduction of Medicare reimbursement for oral
therapies in 2006, and the rollout of a comprehensive Phase IV development
plan during 2005 and beyond, we continue to believe that Tarceva will
emerge as a fundamentally important new medicine in the treatment of
cancer patients around the world.
Moving beyond Tarceva, I would like to comment briefly on some recent
developments in our pipeline before opening up the line for questions. On
the oncology side we have filed a IND for OSI-930, our co-inhibitor of the
c-kit and KDR receptor tyrosine kinases, and receive clearance from the
FDA to begin clinical studies. We anticipate initiating a Phase I trial in
healthy volunteers over the next several weeks. We also advanced a second
candidate, OSI-817, from our Kit/KDR program into IND track development.
We anticipate moving both candidates through the early stages of
development before selecting the better of the two candidates for full
clinical development in cancer patients.
Our U.K. based diabetes and obesity subsidiary, Prosidion, also announced
the initiation of a 60-patient proof-of-concept and dose ranging Phase II
study for PSN9301, our Dipeptidyl Peptidase inhibitor that we acquired
from Probiodrug last year. Dipeptidyl Peptidase-IV cleaves and inactivates
GLP-1, and inhibitors of this enzyme, which are designed to sustain GLP-1
levels in the post-prandial period, are being competitively developed by
many pharmaceutical companies. The most advanced of these is Novartis, who
are in Phase III trials. We expect data from our Phase II study toward the
end of this year.
The glucokinase activated PSN105, and the glycogen phosphorylase inhibitor
PSN357 remain on track to enter clinical trials during the first half of
this year. The group also advanced a second candidate, PSN010, from its
glucokinase activators program onto IND track development.
In summary then, we've had a busy and productive transition quarter. We
are pleased with the launch of Tarceva to date, and we remain focused on
executing on our Tarceva program. As a result of changing market dynamics,
we have been able to bring forward our goal of delivering quarter by
quarter profitability by about a year, to the end of 2006, and are pleased
to report progress in both our oncology and diabetes pipelines.
[Operator provides instructions as to how questions may be asked.]
[Management responds to questions and concludes the webcast conference
call.]