We expect that our revenues from Tarceva and our DPIV patent estate, combined with our diligent management of expenses, will continue to provide us with the capital resources necessary to make disciplined investments in R&D, in order to support the continued growth of Tarceva and our internal pipeline of clinical and pre-clinical assets. As part of our lifecycle plan for Tarceva, we, together with Genentech and Roche, continue to invest in a broad clinical development program directed at maximizing Tarceva's long-term potential, including a number of large, randomized clinical trials designed to expand Tarceva's use in NSCLC. We have also prioritized investment in a portfolio of potentially differentiated and competitive drug candidates and technologies in oncology and diabetes and obesity. As a result, we have successfully advanced four drug candidates into clinical trials over the past two years, all of which were the result of our internal discovery efforts.
In oncology, we have an emerging pipeline of MTTs in clinical and late-stage pre-clinical development which we intend to develop and commercialize independently. These include OSI-906 (an inhibitor of the insulin-like growth factor 1 receptor, or IGF-1R, with potential utility for the treatment of most major solid tumor types), which entered Phase I open-label dose escalation studies in June 2007; OSI-027 (a next generation mammalian target of rapamycin, or mTOR, kinase inhibitor), which entered Phase I studies in July 2008; and OSI-296 (a novel, potent tyrosine kinase inhibitor, or TKI, developed as an epithelial-to-mesenchymal transition, or EMT, inhibitor), which is in late-stage pre-clinical development. Two later stage studies for OSI-906 were commenced in the third quarter of 2009: a Phase III study evaluating the use of OSI-906 for patients with locally advanced or metastatic adrenocortical carcinoma, and a Phase I/II study evaluating OSI-906 in combination with chemotherapy in ovarian cancer patients. These studies are more fully described in the Quarterly Update section below. In addition, we have two anti-angiogenesis agents, OSI-930 and OSI-632, for which we are seeking development partners. In October 2009, we entered into an agreement granting rights to Simcere Pharmaceutical Co., Ltd., a Chinese pharmaceutical company, to develop, manufacture and market OSI-930 in China. Each of these MTTs, as well as Tarceva, are small molecules designed to be administered orally rather than by the less convenient intravenous infusion methods characteristic of most anti-cancer drugs. The focus of our proprietary oncology research efforts is on understanding multiple elements of tumor biology
- including the dependence of certain tumor cells on activated oncogenic signaling pathways, or onco-addiction, and compensatory signaling - but with a particular focus on the biological process EMT which is of emerging significance in understanding tumor development and disease progression. This research has grown out of our translational research efforts to understand which patients may optimally benefit from Tarceva. Our EMT research investment, together with related insights into mechanisms such as compensatory signaling, is the cornerstone of our personalized medicine approach in oncology, and should allow us to better design combinations of MTTs for specific sub-sets of cancer patients. This, in turn, may enable us to realize significant improvements in patient outcomes and to enhance our competitive position in the oncology marketplace. We also have R&D programs in diabetes and obesity which are conducted through Prosidion. Our discovery efforts in diabetes and obesity are concentrated around the neuroendocrine control of bodyweight and glycemia, which focuses on central or peripheral nervous system or hormonal approaches to the control of bodyweight for the treatment of obesity, as well as the lowering of blood glucose together with meaningful weight loss for the treatment of type 2 diabetes. Two compounds from our diabetes and obesity research efforts, PSN821 and PSN602,

entered clinical trials in 2008. PSN821 is an orally administered G protein-coupled receptor 119, or GPR119, agonist with potential anti-diabetic and appetite suppressing features, and PSN602 is an oral dual serotonin and noradrenaline reuptake inhibitor and 5-HT1A agonist for the treatment of obesity.
Quarterly Update
On July 7, 2009, we announced our plans to consolidate our U.S. operations onto a single campus located in Ardsley, New York in Westchester County. On July 20, 2009, we completed the purchase of the 43-acre site, which consists of approximately 400,000 square feet of existing office and laboratory space, for $27 million and expect to incur approximately $60 million to $70 million of capital-related renovation costs over the next fifteen months. In addition, we expect to incur approximately $25 million to $30 million in restructuring-related costs over the next two years, which primarily relate to labor-related and relocation-related costs. We also expect to recognize additional charges related to existing leased facilities when these facilities cease to be used. We will continue to operate our diabetes/obesity franchise in Oxford, England.
On July 13, 2009, we, along with Genentech, announced that the SATURN study met a key secondary endpoint of extending overall survival in patients with advanced NSCLC when Tarceva was used as single agent, maintenance therapy in patients who did not progress following first-line treatment with platinum-based chemotherapy. On August 1, 2009, we announced further results from the SATURN study, as presented at the 13th World Conference on Lung Cancer held in San Francisco. The study showed that patients with NSCLC treated with Tarceva had a 23% improvement in overall survival compared with patients who received placebo (hazard ratio=0.81; p-value=0.0088; a hazard ratio of less than one for survival indicates a reduced risk of death). The hazard ratio, which assesses risk in the overall trial population, is widely recognized as the best measure of overall benefit in large randomized clinical trials. The median survival (a single point estimate of benefit) for patients receiving Tarceva was 12 months versus a median survival of 11 months for patients receiving placebo. The study confirmed that a broad spectrum of patients with advanced NSCLC experienced a survival benefit from Tarceva. Specific analysis of patients in the study whose tumors were confirmed to be EGFR "wild-type" - i.e., not having an EGFR activating mutation - experienced a 30% improvement in survival (hazard ratio=0.77; p-value=0.0243). The majority of patients with NSCLC are EGFR wild-type. Overall survival for the patient sub-group with EGFR mutations is still immature with the median survival not yet being reached in patients with EGFR mutations receiving Tarceva. Determination of overall survival in this sub-group has been confounded by the fact that two-thirds of the patients with EGFR mutations who received placebo rather than Tarceva then crossed over to receive Tarceva or another EGFR therapy. We believe that, assuming it is approved for use as maintenance therapy following initial chemotherapy, Tarceva will provide a new, convenient, non-chemotherapy treatment option for patients without exposing them to the continuous burden and lifestyle constraints of long-term chemotherapy.
On July 21, 2009, we, together with AVEO Pharmaceuticals, Inc., or AVEO, announced that we had expanded our existing drug discovery and translational research collaboration. As part of the expanded collaboration, we made a $5 million upfront payment to AVEO and purchased an additional $15 million of AVEO preferred equity. The alliance is anchored around developing molecular targeted therapies to target the underlying mechanisms of EMT in cancer and to develop proprietary datasets of associated patient selection biomarkers to support our tar-

geted medicine pipeline. Together, we and AVEO are expanding our efforts to validate cancer targets and to deploy key elements of AVEO's proprietary translational research platform in support of our clinical development programs.
On September 3, 2009, we announced the initiation of two clinical trials with OSI-906, our IGF-1R inhibitor. The first study is a Phase III, multi-center study that will evaluate the use of OSI-906 for patients with locally advanced or metastatic adrenocortical carcinoma. The study is designed to determine overall survival for patients receiving single-agent OSI-906 versus placebo and will also evaluate progression free-survival, disease control rate, overall response rate as well as safety. The second study is a Phase I/II trial evaluating OSI-906 in combination with the chemotherapy Taxol® (paclitaxel) primarily in patients with recurrent epithelial ovarian cancer. In addition, we intend to initiate a registration-oriented combination study of OSI-906 with Tarceva in NSCLC. Initial indications of monotherapy activity for OSI-906 in NSCLC and our translational research data suggesting that compensatory signaling mechanisms and EMT phenomena may make a combination of Tarceva and OSI-906 synergistically effective in the NSCLC setting provide the rationale for this clinical trial, which is intended to begin in 2010.
On September 17, 2009, we announced that the U.S. Patent & Trademark Office issued a "Notice of Allowance" in our reissue application for U.S. Patent No. 5,747,498, the composition of matter patent for Tarceva, or the '498 patent. In February 2008, we had filed an application to reissue the '498 patent in order to correct certain errors relating to the claiming of compounds, other than Tarceva, which fall outside of the scope of the main claim in the patent. The reissue application looked to correct these errors by deleting surplus compounds from the claims. Like most composition of matter patents, the '498 patent claims many compounds in addition to Tarceva. Tarceva itself is accurately described in the '498 patent. The reissued patent will replace the original '498 patent and will have the same November 2018 expiration date. At the time of reissue, a new patent number will be assigned. The reissued patent will include new claims that solely identify Tarceva. Subsequent Events
In October 2009, Roche communicated to us that an exploratory data sweep for survival of ATLAS, a Phase III study in patients with advanced NSCLC who received Tarceva in combination with Avastin as first-line maintenance therapy, was not positive. Survival was a secondary endpoint in the ATLAS study and the study was not powered to detect a significant difference in overall survival. We did not "opt-in" to the ATLAS study because the study was not designed to be registrational. If the study were to result in an ATLAS-related change to the Tarceva label we would likely be required to opt-in and pay our retrospective share of the study costs and a penalty. We are not currently forecasting any opt-in payments for this study.
Critical Accounting Policies
For a discussion of our critical accounting policies, see "Item 7. Management's Discussion and Analysis of Financial Condition and Results of Operations" in our annual report on Form 10-K for the fiscal year ended December 31, 2008. There have been no material changes to our critical accounting policies for the nine month period ended September 30, 2009.

Revenues

                                             Three Months Ended                                      Nine Months Ended
                                                September 30,                                          September 30,
(In thousands)                    2009               2008            $ Change            2009               2008             $ Change

Tarceva-related revenues       $  88,735          $ 80,708          $  8,027          $ 257,914          $ 251,006          $  6,908
Other revenues                    22,712            13,864             8,848             46,276             29,955            16,321

Total revenues                 $ 111,447          $ 94,572          $ 16,875          $ 304,190          $ 280,961          $ 23,229

Tarceva-Related Revenues
Tarceva-related revenues for the three and nine months ended September 30, 2009 were $88.7 million and $257.9 million, respectively, compared to Tarceva-related revenues for the three and nine months ended September 30, 2008 which were $80.7 million and $251 million, respectively. Tarceva-related revenues include net revenue from our unconsolidated joint business, Tarceva-related royalties and Tarceva-related milestones.
Net Revenue from Unconsolidated Joint Business Net revenue from unconsolidated joint business is related to our co-promotion and manufacturing agreements with Genentech for Tarceva. For the three and nine months ended September 30, 2009, Genentech recorded net sales of Tarceva in the United States and its territories of approximately $118 million and $342 million, respectively, compared to approximately $110 million and $340 million for the three and nine months ended September 30, 2008, respectively. The increase in net sales of Tarceva for the three months ended September 30, 2009 compared to the same period last year was primarily a result of price increases and a lower level of reserve adjustments in 2009, partially offset by lower unit volume sales. The increase in net sales of Tarceva for the nine months ended September 30, 2009 compared to the same period last year was primarily a result of price increases, partially offset by a higher level of reserve adjustments in 2009 and by lower unit volume sales. On a sequential quarter-over-quarter basis, sales of Tarceva for the three months ended September 30, 2009 increased $5.3 million from approximately $113 million for the three months ended June 30, 2009, primarily as a result of a lower level of reserve adjustments recorded for the three months ended September 30, 2009 and a price increase that occurred in September 2009.
Our share of Tarceva net sales is reduced by the costs incurred for cost of goods sold and for the sales and marketing of the product. For the three and nine months ended September 30, 2009, we reported net revenues from our unconsolidated joint business for Tarceva of $50.8 million and $148.5 million, respectively, compared to $45.5 million and $147.2 million for the three and nine months ended September 30, 2008, respectively. The increase in net revenue from unconsolidated joint business for the three months ended September 30, 2009 was primarily due to higher net sales, higher reimbursement of our marketing and sales costs and lower overall sales and marketing costs incurred by the collaboration. The increase in net revenue from unconsolidated joint business for the nine months ended September 30, 2009 was primarily due to higher net sales and higher reimbursement of our marketing and sales costs, partially offset by higher overall sales and marketing costs incurred by the collaboration.

Tarceva-Related Royalties
We receive royalties from Roche of approximately 20% on net sales of Tarceva outside of the United States and its territories. The royalty amount is calculated by converting the respective countries' Tarceva sales in local currency to Roche's functional currency (Swiss francs) and then to U.S. dollars. The royalties are paid to us in U.S. dollars on a quarterly basis. As a result, fluctuations in the value of the U.S. dollar against the Swiss franc and local currencies in which Tarceva is sold will impact our earnings. For the three and nine months ended September 30, 2009, Roche reported U.S. dollar equivalent rest of world sales of approximately $183 million and $528 million, respectively, compared to approximately $169 million and $498 million in the same periods last year, respectively. For the three and nine months ended September 30, 2009, we recorded $36.9 million and $106.6 million in royalty revenue from these sales, respectively, compared to $34.2 million and $100.8 million in the same periods last year, respectively. The increase in royalty revenue for the three and nine months ended September 30, 2009 was primarily due to increased sales volume outside the United States, partially offset by the negative impact of net unfavorable changes in foreign exchange rates.
Tarceva-Related Milestones
Milestone revenue from Tarceva includes the recognition of the ratable portion of upfront fees from Genentech and milestone payments received from Genentech and Roche to date in connection with various regulatory acceptances and approvals for Tarceva in the United States, Europe and Japan. These payments were initially deferred and are being recognized as revenue ratably over the term of the agreement. The ratable portions of the upfront fee and milestone payments recognized as revenue for the three and nine months ended September 30, 2009 were $921,000 and $2.8 million, respectively, compared to $979,000 and $2.9 million for the same periods last year, respectively. The unrecognized deferred revenue related to these upfront fees and milestone payments received was $34.3 million and $37.1 million as of September 30, 2009 and December 31, 2008, respectively. We also will be entitled to additional milestone payments from Genentech and Roche upon the occurrence of certain regulatory approvals and filings with respect to Tarceva. Additional milestone payments will be due from Genentech and Roche upon approval of adjuvant indications in the United States and Europe. The ultimate receipt of these additional milestone payments is contingent upon the applicable regulatory approvals and other future events. Other Revenues
Other revenues for the three and nine months ended September 30, 2009 were $22.7 million and $46.3 million, respectively, compared to $13.9 million and $30.0 million for the same periods last year, respectively, and include non-Tarceva related license, milestone, royalty and commission revenues.
We recognized $17.3 million and $40.1 million of royalty revenue for the three and nine months ended September 30, 2009, respectively, compared to $12.8 million and $26.8 million for the same periods last year, respectively, from previously granted worldwide non-exclusive license agreements entered into by Prosidion under our DPIV patent estate covering the use of DPIV inhibitors for treatment of type 2 diabetes and related indications. Our royalty revenue for

the three and nine months ended September, 2009 and 2008 was principally derived from sales of Merck's DPIV inhibitor product, Januvia ® (sitagliptin), and its combination product, Janumet®(sitagliptin/metformin). We also derived royalty revenue from sales of Novartis AG's DPIV inhibitor product, Galvus ® (vildagliptin), and its combination product Eucreas ® (vildagliptin/metformin), and Bristol-Myers Squibb's DPIV inhibitor Onglyza™ (saxagliptin).
Other revenues for the three and nine months ended September 30, 2009 also include a $5 million milestone payment related to a non-exclusive licensing agreement under our DPIV patent portfolio.

Expenses

                                           Three Months Ended                                     Nine Months Ended
                                             September 30,                                          September 30,
(In thousands)                  2009              2008            $ Change            2009               2008             $ Change

Cost of goods sold           $  1,607          $  2,517          $   (910 )        $   6,460          $   6,748          $   (288 )
Research and
development                    38,546            33,054             5,492            111,129             94,009            17,120
Acquired in-process
research and
development                     5,000                 -             5,000              5,000                  -             5,000
Selling, general and
administrative                 24,677            22,262             2,415             74,065             69,985             4,080
Restructuring costs             1,148                 -             1,148              1,148                  -             1,148
Amortization of
intangibles                       229               646              (417 )              693              1,884            (1,191 )

                             $ 71,207          $ 58,479          $ 12,728          $ 198,495          $ 172,626          $ 25,869

Cost of Goods Sold
Total cost of goods sold for the three and nine months ended September 30, 2009 was $1.6 million and $6.5 million, respectively, compared to $2.5 million and $6.7 million for the three and nine months ended September 30, 2008, respectively, and represents the cost of goods sold related to Tarceva. Research and Development
R&D expenses for the three and nine months ended September 30, 2009 increased $5.5 million and $17.1 million, respectively, compared to the same periods last year. The increase was primarily due to an increase in R&D expenses related to non-Tarceva oncology programs, in particular OSI-906, our IGF-1R inhibitor candidate and our diabetes and obesity programs, and equity-based compensation, partially offset by a decline in R&D expenses for Tarceva.
We manage the ongoing development program for Tarceva with our collaborators, Genentech and Roche, through a global development committee under a Tripartite Agreement among the parties. Together with our collaborators, we have implemented a broad-based global development strategy for Tarceva that implements simultaneous clinical programs currently designed to expand the number of approved indications for Tarceva and evaluate the use of Tarceva in new and/or novel combinations. Since 2001, we and our collaborators have committed an aggregate of approximately $930 million to the global development plan to be shared by the three parties. As of September 30, 2009, we had invested in excess of $266 million in the development of Tarceva, representing our share of the costs incurred through September 30, 2009 under the tripartite global development plan and additional investments outside of the plan.

We consider the active management and development of our clinical pipeline crucial to the long-term process of getting a clinical candidate approved by the regulatory authorities and brought to market. We manage our overall research, development and in-licensing efforts in a highly disciplined manner designed to advance only high quality, differentiated agents into clinical development. The duration of each phase of clinical development and the probabilities of success for approval of drug candidates entering clinical development will be impacted by a variety of factors, including the quality of the molecule, the validity of the target and disease indication, early clinical data, investment in the program, competition and commercial viability. Because we manage our pipeline in a dynamic and disciplined manner, it is difficult to give accurate guidance on the anticipated proportion of our R&D investments assigned to any one program prior to the Phase III stage of development, or to the future cash inflows from these programs. For the three and nine months ended September 30, 2009, we invested a total of $14.1 million and $40.0 million, respectively, in research, and $24.4 million and $71.1 million, respectively, in pre-clinical and clinical development. For the three and nine months ended September 30, 2008, we invested a total of $13.0 million and $38.3 million, respectively, in research, and $20.1 million and $55.7 million, respectively, in pre-clinical and clinical development. We believe that this represents an appropriate level of investment in R&D for our company when balanced against our goals of financial performance and the creation of longer-term shareholder value. Acquired In-process Research and Development In July 2009, we expanded our drug discovery and translational research . . .