Back to GetFilings.com

[X] ANNUAL REPORT PURSUANT TO SECTION 13 OR 15(d)
OF THE SECURITIES EXCHANGE ACT OF 1934
For the Fiscal Year Ended December 31, 2003
OR
[   ] TRANSITION REPORT PURSUANT TO SECTION 13 OR 15(d)
OF THE SECURITIES EXCHANGE ACT OF 1934

Commission File Number 0-18311

NEUROGEN CORPORATION
(Exact name of registrant as specified in its charter)

(203) 488-8201
(Registrant's telephone number, including area code)

Securities registered pursuant to Section 12(b) of the Act:
None

Securities registered pursuant to Section 12(g) of the Act:
Common Stock, par value $.025 per share (the "Common Stock")
(Title of Class)

Indicate by check mark whether the registrant (1) has filed all reports required to be filed by Section 13 or 15(d) of the Securities Exchange Act of 1934 during the preceding 12 months (or for such shorter period that the registrant was required to file such reports), and (2) has been subject to filing requirements for the past 90 days. YES [X] NO [   ]

Indicate by check mark if disclosure of delinquent filers pursuant to Item 405 of Regulation S-K is not contained herein, and will not be contained, to the best of registrant’s knowledge, in definitive proxy or information statements incorporated by reference in Part III of this Form 10-K or any amendment to this Form 10-K. [X]

Indicate by check mark whether the registrant is an accelerated filer (as defined in Rule 12b-2 of the Act). YES [X] NO [    ]

The approximate aggregate market value of the registrant’s Common Stock held by non-affiliates was $65,622,906, based on the closing price of a share of Common Stock as reported on the NASDAQ National Market on June 30, 2003, which is the last business day of the registrant’s most recently completed second fiscal quarter. In determining the market value of non-affiliate voting stock, shares of Common Stock beneficially owned by officers and directors and one shareholder owning more than 20% of the Common Stock have been excluded from the computation. This number is provided only for purposes of this report and does not represent an admission by either the registrant or any person as to the status of such person.

As of March 1, 2004, the registrant had 19,892,646 shares of Common Stock outstanding.

DOCUMENTS INCORPORATED BY REFERENCE

(1) Portions of the Neurogen Corporation Proxy Statement for the Annual Meeting of Stockholders to be held on May 21, 2004 are incorporated by reference into Items 10, 11, 12, 13 and 14 of Part III of this Form 10-K.

NEUROGEN CORPORATION
ANNUAL REPORT ON FORM 10-K

For the Year Ended December 21, 2003

TABLE OF CONTENTS

SIGNATURES

EXHIBIT INDEX

PART I

ITEM 1. BUSINESS

Overview

Neurogen Corporation (Nasdaq: NRGN) ("Neurogen" or the "Company"), incorporated under the laws of the State of Delaware in 1987, is a drug discovery and development company targeting new small molecule drugs to improve the lives of patients suffering from disorders with significant unmet medical needs including inflammation, pain, insomnia, depression and obesity. Neurogen has generated a portfolio of compelling new drug programs through its fully integrated drug discovery and development processes. Neurogen’s Accelerated Intelligent Drug Discovery ("AIDD" ™) process and its expertise in cellular functional assays enhance the Company's ability to rapidly and cost effectively identify small molecule drug candidates. Small molecule drugs typically are suitable for oral administration as a pill, while large molecule drugs typically are administered by injection. The Company conducts its research and development independently and, when advantageous, collaborates with leading pharmaceutical companies during the drug research and development process to obtain additional resources and to access complementary expertise. Currently, Neurogen has active collaborations with Pfizer Inc. ("Pfizer"), Aventis Pharmaceuticals, Inc. ("Aventis"), and Merck Sharp & Dohme Limited ("Merck"), a subsidiary of Merck & Co., Inc.

Neurogen has diversified its drug discovery and development efforts across a broad number of disease-related targets and employs a strategy designed to efficiently discover multiple drug candidates for each target. In addition, several of the Company’s programs focus on medical targets which are implicated in numerous, related disease states. Throughout the pharmaceutical industry, a small minority of all drug candidates successfully overcome all of the development obstacles on the way to commercialization. Because of this very high attrition rate, the Company believes that the value of a drug discovery and development company’s pipeline is best measured by the company’s ability to rapidly discover multiple generations of improved candidates within each of several programs.

Neurogen believes that its ability to rapidly and systematically produce multiple generations of incrementally improved drug candidates in multiple programs represents a competitive advantage. Although much smaller than major pharmaceutical companies, Neurogen has discovered ten small molecule drug candidates that, in collaboration with our pharmaceutical company partners, have entered human clinical trials. During 2003, the Company completed an exploratory Phase IIa trial (see Government Regulation section below) in asthma of its lead drug candidate for the treatment of inflammatory disorders (see Neurogen Drug Programs section below). An exploratory Phase IIa study in rheumatoid arthritis, also begun during 2003, continued into 2004. Neurogen owns all commercial rights to this program. Our partner, Pfizer, continued to conduct Phase I human trials, begun in March 2002, for a drug candidate from the Company's program for the treatment of insomnia.

Background on the Drug Discovery Industry

Most drugs work by binding to a particular protein, cell, or receptor, called a target, in the human body, thereby altering communication between cells or otherwise regulating cellular activity. Therefore, the traditional path to discovering small molecule drugs typically begins with the identification of a biological target that is believed to regulate cellular communication or activities, which could be modulated to treat a given disorder. A test, or assay, is then developed in order to expose many chemical compounds to the biological target and determine whether any of the compounds interact with the target. Such an assay facilitates the screening of the target against a library of many compounds that have been synthesized in the laboratory. Compounds that bind to the target protein and alter its activity are referred to as "hits." Medicinal chemists then create many new compounds that are chemically similar to the original hit in an effort to improve, or optimize, these newer hits until they have sufficient potency to become lead drug candidates and then improve their "drug-like" properties, such as gastrointestinal absorption, stability, freedom from unwanted activities, etc., with the goal of producing a successful drug development candidate.

Some drug companies try to streamline the drug discovery process by copying chemical structures from known active compounds. Even taking this approach, however, the number of possible compounds that could be made is too vast to actually test against even a single target using any available technology. Generally, the search is further narrowed only by educated guessing. As a result of the uncertainty of this approach, traditional methods can take many years or may fail entirely.

If it were possible to predict in advance which compounds would result in a hit, and which chemical changes would help optimize hits into drug candidates, the drug discovery process would be vastly simplified. Unfortunately, the traditional drug discovery process has had to rely on a trial and error approach that has proven extremely expensive, inefficient and unreliable. Optimization of hits to achieve the extremely delicate balance of properties necessary for a successful drug is still a daunting task. Most hits are never optimized into successful drugs despite years of effort.

Neurogen’s Competitive Advantage

Neurogen’s drug discovery and development platform is designed to rapidly discover small molecule drug candidates for medical targets where other pharmaceutical or biotechnology companies have failed or for recently discovered targets. The Company’s proprietary AIDD™ platform enables the rapid and efficient discovery of compounds that hit potential drug targets, evaluate the utility of those targets, and optimize useful hits into new drug candidates. Neurogen’s VR1 program for the management of pain, now partnered with Merck, is an example of the Company’s ability to establish a leading position in an important new area of drug discovery. Neurogen’s ability to rapidly discover small molecule drug candidates enabled it to quickly generate hits sufficient to evaluate the potential utility of this new target for the relief of pain and then establish a significant intellectual property position. In addition to establishing the first patent publications on small molecule VR1 antagonists, Neurogen has been awarded a patent on a human VR1 receptor gene sequence, has filed broad applications encompassing highly potent chemotypes, and has patent applications pending on the general use of VR1 antagonists for the treatment of pain.

Neurogen discovers and develops small molecule drugs. Small molecule drugs are usually more stable and easily absorbed than large molecule drugs, and so in most cases may be administered as an orally-administered pill, a patch, or an ointment. In addition, small molecule drugs are generally much easier and less expensive to manufacture, distribute, and store. Protein-based large molecule drugs typically are administered by injection and require refrigeration, while most small molecule drugs do not. Small molecule drugs can also be safely shipped and stored at regular temperatures. Small molecule drugs that can be taken orally currently make up about three quarters of the sales of the top 100 prescription drugs. Where there is a choice, patients generally would rather take a pill by the oral route of administration than an injection.

Neurogen's Business Strategy

Neurogen’s goal is to bring approved drugs to patients suffering from disorders with unmet medical need. Key elements of the Company’s business strategy to achieve this goal follow:

• Create a risk-balanced drug portfolio. To increase the probability of successful drug discovery and development efforts, Neurogen is pursuing multiple targets with multiple drug candidates. The Company is concentrating its efforts on key therapeutic areas (inflammation, pain, insomnia, depression and obesity) within which clinical indications may be developed for a variety of disorders. The Company believes that its ability to efficiently produce multiple generations of candidates in multiple programs using AIDD™ represents an important element of its strategy to address the very high attrition rates inherent in the drug development process and increase incrementally the probability of successfully advancing one or more candidates to market.
  


• Expand development capability. Neurogen possesses a powerful drug discovery capability, which is closely linked to pharmaceutical development and clinical initiatives to continuously optimize drug properties. The Company believes this integrated approach will aid in the development of compounds in the clinic, and is deploying additional resources to support the early clinical trials of lead drug candidates.
  


• Selectively partner drug programs. Neurogen's partnering strategy seeks to retain ownership and control of some programs – taking them further into human clinical testing – while partnering other programs at earlier stages, so that the Company’s risk is balanced while seeking to maximize return for shareholders. Neurogen’s development capabilities enable the Company to pursue a flexible business model of partnering programs when competitively and economically advantageous. When Neurogen partners programs, it seeks to collaborate with pharmaceutical leaders with a complementary set of demonstrated strengths and resources.
  

Components of Neurogen's Discovery and Development Platform

The Company’s AIDD™ system, as well as its focused compound library and Virtual Library™, are key components of the drug discovery infrastructure. AIDD™ is an integrated system of hardware, software, and processes that allow scientists to improve on the trial and error approach traditionally associated with drug discovery and development. This system incorporates automated robotics guided by state-of-the-art computerization, including neural network-based artificial intelligence, to aid the Company’s scientists in the design, modeling, synthesis and screening of new chemical compounds.

Neurogen’s AIDD™ system works in tandem with the Company’s focused compound library. Instead of randomly generating a compound library as many other pharmaceutical and drug discovery companies have done, Neurogen has chosen to bias or "enrich" its compound library in favor of selected families of compounds. Because the number of small molecule organic compounds that can be synthesized is virtually infinite, the Company believes that to be successful in the drug discovery process, it is not the biggest or most diverse library that counts, but rather the richest and most intelligently designed.

Also critical to Neurogen’s drug discovery and development capabilities is the Company’s biological expertise. Neurogen believes that its expertise in receptor biology is a competitive advantage and utilizes this expertise in the design and construction of screening assays to capitalize on medical targets. The Company focuses on discovering novel drugs, and all of the drug candidates that have entered human testing in clinical trials have worked by distinctly new target mechanisms, which have the potential to provide significant therapeutic advantages over existing drugs.

Neurogen Drug Programs

By discovering and developing drug candidates specifically targeting a new mechanism, Neurogen’s programs offer the potential for introducing new drugs which are differentiated from drugs currently on the market. In addition, drugs active against new targets may offer the possibility of adding to or synergistically improving existing therapy.

The Company is in the early stage of product development and does not expect to have any products resulting from its research efforts commercially available for a number of years, if at all. To date, the Company has stopped developing eight compounds which had previously been in clinical development. Currently, Neurogen has several compounds in preclinical development, one compound in Phase IIa human clinical trials and one compound in Phase I human clinical trials. All of the Company's compounds currently being developed, whether in human clinical trials or not will require significant additional research, development and testing before they can be commercialized. Neurogen cannot accurately predict the time required or the cost involved in commercializing any new drug. In addition, developing new drugs is an extremely uncertain process, and unanticipated developments such as clinical or regulatory delays, unexpected side effects in test patients, or ineffectiveness would slow or prevent the development of a product. If Neurogen is unable to commercialize any drug products, the Company will never achieve product revenues and may eventually be unable to continue operating. This result would cause investors to lose all or a substantial portion of their investment.

The following table summarizes the most advanced programs in the Company’s current drug pipeline:

Below, Neurogen’s most significant active drug development programs are described in detail.

Inflammation Program: C5a receptor antagonist

Inflammatory and autoimmune conditions are debilitating diseases afflicting millions of Americans, causing substantial human suffering, and impacting the United States economy with significant health care costs and lost productivity. These diseases include many different forms of arthritis, autoimmune diseases, and pulmonary disorders. Most are chronic and many cause life-long pain, disability, or disfigurement. These diseases affect people of all ages, racial and ethnic populations, and economic strata.

The body’s natural immune system includes a biological response referred to as the complement cascade. One component of the complement system is represented by the C5a protein, which promotes inflammation, attracts white blood cells, and may trigger the immune system to start attacking the body's own tissues, an inappropriate reaction central to autoimmune and inflammatory diseases. Neurogen scientists believe that blocking the activation of the C5a receptor of the immune system’s complement cascade may work to treat inflammation in rheumatoid arthritis (RA) and other autoimmune diseases.

RA involves inflammation in the lining of the joints and/or other internal organs. It is a systemic disease affecting the entire body and is one of the most common forms of arthritis, characterized by the inflammation of the membrane lining the joint, resulting in pain, stiffness, warmth, redness and swelling. Industry surveys indicate that this disease affects 2.6 million Americans.

Certain types of arthritis can occur as early as infancy. In children, the most common form is called juvenile rheumatoid arthritis ("JRA"). The Arthritis Foundation estimates that this condition affects an estimated 30,000 to 50,000 children in the Unites States. The American College of Rheumatology estimates that the number of children in the U.S. suffering from JRA may be as high as 150,000.

Currently, several types of drugs are used to treat rheumatoid arthritis. Recently, drugs which inhibit the tumor necrosis factor alpha ("TNFa") protein, such as Enbrel® and Remicade®, have demonstrated the ability to not only treat the symptoms of rheumatoid arthritis but also slow the progression of the disease by inhibiting the destruction of healthy joint tissue. However, as large molecule proteins, these drugs must be taken by injection and are expensive to manufacture, distribute and administer. Anti-tumor necrosis factor drugs also have some side effects such as development of antinuclear antibodies (these are antibodies seen in various autoimmune diseases such as Lupus); Urticaria (reddish blotchy areas on the skin), infections, and rejection of these antibodies by the patient’s own immune system (similar to the body's rejection of an organ transplant).

Other drugs used to treat rheumatoid arthritis include the COX-2 inhibitors. Since the synthesis of aspirin in 1897, inhibitors of cyclooxygenase (an enzyme that controls inflammation) have been a source of relief from arthritis pain. These drugs are now collectively called non-steroidal anti-inflammatory drugs ("NSAID’s"). The newest generation of NSAID’s is called COX-2 selective inhibitors. Brand names of this type of medication include Celebrex® and Vioxx®. While COX-2 inhibitors are small molecule drugs and feature less risk of side effects such as blood thinning or stomach ulceration than earlier treatments, these drugs do not slow the tissue destruction associated with the progression of rheumatoid arthritis.

Neurogen believes that blocking the C5a receptor in patients with rheumatoid arthritis may inhibit the destruction of healthy tissue by blunting the immune system’s overreaction to the presence of the C5a protein. The Company believes that its small molecule drug candidate, taken as a pill for the treatment of rheumatoid arthritis, may offer many of the advantages of the COX-2 inhibitors with the potential, because a small molecule drug should be able to enter the synovial fluid of joints, to also slow the progression of the disease.

The Company has identified several small molecule compounds that potently block the activation of C5a receptors. Phase I study results indicated that the Company’s lead compound in this program, NGD 2000-1, was safe and well tolerated in single and multiple ascending doses and also identified that the compound inhibits cytrochrome P450 3A4 ("3A4"), a metabolic enzyme found in the liver, which metabolizes many drugs. This characteristic may limit the use of NGD 2000-1 by patients also taking other medications which are metabolized by the 3A4 enzyme. Further studies will be required to determine the extent of any such limitation, to development or commercialization.

NGD 2000-1 is currently in exploratory Phase IIa clinical trials for treatment of RA. In this exploratory study, Neurogen is examining the ability of the compound to lower the levels of C-reactive protein ("CRP"), a biomarker in RA, in patients with mild to moderate conditions. The study is also examining secondary endpoints of clinical signs and symptoms. The study is multi-centered, double-blinded and placebo-controlled and explores multiple doses of NGD 2000-1. Neurogen expects to conclude this study in the second quarter of 2004. The Company will evaluate its development plan for NGD 2000-1 based on the results of the Phase IIa study and further testing regarding the 3A4 enzyme inhibition, designed to identify the extent of specific drug interactions. NGD 2000-1 may be more appropriate for use in patient populations where the number of medications prescribed per patient is low. JRA may be an example of a patient population with few concomitant medications and where NGD 2000-1 might offer therapeutic benefit. It is also possible that the Company may determine that NGD 2000-1 is not appropriate for further development.

In January 2004, results of an exploratory Phase IIa trial studying the safety and preliminary efficacy of NGD 2000-1 in patients with mild to moderate asthma were reported. The trial was multi-centered, double-blinded, and placebo-controlled and assessed multiple doses of NGD 2000-1 using improvement in lung capacity as measured by forced expiratory volume in one second (FEV-1) as the primary endpoint. In this study, NGD 2000-1 did not demonstrate a statistically significant therapeutic benefit with respect to the primary and most secondary predetermined endpoints.

To date, Neurogen has retained all rights to its C5a antagonist program.

Insomnia Program: unique GABA modulation profile

Recent studies indicate that as many as 20 million people in the United States experience chronic insomnia and an additional 20 to 30 million Americans experience intermittent sleep disorders. While currently marketed drugs to treat sleep disorders, known as hypnotics, are effective, they may cause numerous side effects, including "hangovers," rebound insomnia, short-term memory loss and addiction. Neurogen is developing drugs to treat sleep disorders, primarily insomnia, which selectively modulate certain subtypes of the gamma-aminobutyric acid ("GABA") neurotransmitter system.

The link between the GABA system and sleep is illustrated by the Benzodiazepine class of drugs such as Valium®, which cause sleepiness, and by drugs marketed to treat insomnia such as Ambien® and Sonata®, which work through the same broad range of GABA receptors as the benzodiazepines. Neurogen has identified drug candidates to treat insomnia that have a more specific GABA receptor binding profile than currently marketed drugs. Animal studies, to date, suggest that these compounds are efficacious in inducing sleep with fewer side effects than existing therapies. Drugs to treat insomnia should not only induce sleep but they should have pharmacokinetic properties, which cause the drug to work quickly and then be out of the system before morning.

Pfizer, Neurogen's partner in this program, is currently evaluating the Company's lead compound, NGD 96-3, which entered Phase I clinical studies in the first quarter of 2002. Pfizer has now completed three single dose studies exploring the safety, pharmacokinetics and preliminary efficacy of NGD 96-3 and is further assessing the safety profile and commercial potential of the compound. Neurogen has initiated discussions with Pfizer to explore various options to advance the clinical development of NGD 96-3.

In a Phase I, placebo-controlled, single rising dose study of the safety and pharmacokinetics of NGD 96-3, the compound was well-tolerated in healthy volunteers. Reports of sleepiness by treated subjects in this study were consistent with the expected action of the drug.

Pfizer has also performed a separate human efficacy study of two dose levels of NGD 96-3 in healthy volunteers using the marketed drug zolpidem (Ambien®) and placebo as controls. In that study, when compared to subjects taking a placebo, both doses of NGD 96-3 significantly reduced the latency (or length of time) required for healthy subjects to fall into an EEG-defined sleep state during the day.

To further assess the safety profile of NGD 96-3, Pfizer has also conducted an additional single dose study in healthy volunteers. Given the broad patient population eligible for treatment and the availability of existing therapies, assessing the safety profile is a critical step in the development of drugs for the treatment of insomnia.

Pain Management Program: VR1 receptor antagonist

Industry sources estimate more than 100 million people in the United States suffer from some type of acute or chronic pain sufficient to significantly impact their lives. Not only does such chronic pain adversely affect physical and psychological well being, it also costs society in lost productivity, health care expenditures, and disability compensation. Among cancer patients, available estimates suggest that chronic pain affects up to 95% of patients with advanced disease with less than 50% of these experiencing effective pain relief.

The Company believes that its approach to pain management offers the potential to relieve pain without the side effect profile of existing pain drugs. Neurogen and its partner Merck have advanced lead candidates from their vanilloid receptor-1 program into formal preclinical development and are also advancing additional, highly potent candidates for potential development in this program. The program is initially targeting pain and urinary incontinence, with additional indications possible.

Studies which model inflammatory pain in genetically altered mice lacking the VR1 receptor indicate a role of VR1 role in the sensation of noxious heat as well as thermal hyperalgesia (heightened sensitivity to pain). Neurogen researchers believe that a drug which blocks the VR1 receptor could benefit patients suffering from various types of inflammatory pain states and specific types of nociceptive, or localized, pain.

Neurogen has established a significant intellectual property position on this new target for the relief of pain. In addition to establishing the first patent publications on small molecule VR1 antagonists, the Company has been awarded a patent on a human VR1 receptor gene sequence, has broad applications filed encompassing highly potent chemical templates, and has patent applications pending on the general use of VR1 antagonists for the treatment of pain.

Neurogen entered into a collaboration agreement for VR1 research and development with Merck (described below) in December 2003.

Depression/Anxiety Program: CRF-1 receptor antagonist

Depression and anxiety are two of the most prevalent mental illnesses in the United States, affecting approximately 22 million people or 11.6% and 9 million or 4.8%, respectively, of the adult population annually, according to recent industry surveys. While recent pharmaceutical research has led to improved drugs such as Prozac® for the treatment of depression, these medications are limited in their use, primarily because of slow onset of therapeutic action (often greater than 10 days from the commencement of dosing), lack of efficacy in some patients, and side effects such as sexual dysfunction.

Anxiety and stress are conditions commonly associated with depression. A number of neuropeptide receptors that appear to be involved in stress responses, including receptors for CRF-1, exhibit altered characteristics in depressed patients.

Neurogen believes that an orally available drug candidate that blocks the CRF-1 receptor may be efficacious in relieving depression, anxiety and/or stress related disorders. A number of companies are seeking to develop CRF-1 drug candidates. To date, many companies have experienced difficulties in identifying CRF-1 blockers, which have drug properties appropriate for commercialization. Neurogen believes this is due to the fact that the scope of known chemical structures which block CRF-1 is relatively narrow.

Neurogen has discovered a number of compounds that block the CRF-1 target and have demonstrated efficacy in animal models of depression and stress. Importantly, the chemical structure of these compounds is significantly outside of the scope of known CRF-1 blockers. The Company believes these novel chemical templates may provide an avenue for avoiding many of the development issues of earlier CRF-1 structures investigated in the industry.

In December of 2001, Neurogen entered into a collaborative agreement with Aventis (described below) to research and develop compounds to modulate CRF-1 receptors. Progress in the Company’s research and development collaboration with Aventis resulted in the achievement of a preclinical milestone in December 2003, which triggered the payment of $1 million to Neurogen in January 2004. Neurogen and Aventis are working together to advance their preclinical compound through the rigorous testing necessary prior to an IND filing.

Obesity/Diabetes Program: MCH-1 receptor antagonist

Neurogen is optimizing lead compounds blocking the melanin concentrating hormone receptor-1 ("MCH-1"), which is an important mediator of food intake. These receptors are expressed in the hypothalamus, the feeding center of the brain. While obesity is caused by a complex process involving many hormones, neurotransmitters, nerve cells, and genes, the MCH neurotransmitter is now known to play a key role in controlling eating behavior. In preclinical studies, removing the MCH peptide or receptor gene has resulted in lean animals, while administering MCH caused increased weight gain. The Company is currently optimizing lead drug candidates in this program and further exploring in animal models the potential utility of MCH-1 antagonists. Neurogen owns all commercial rights to its MCH-1 receptor antagonist program.

Neurogen Collaborations

Neurogen conducts its research and development independently and, when advantageous, collaborates with global pharmaceutical companies during the drug development process to obtain additional resources and to access complementary expertise. The Company's collaboration agreements may provide funding for drug discovery and development programs as well as clinical, manufacturing, marketing, and sales expertise. At the same time, Neurogen has retained certain rights to future royalties or profit sharing for successful drugs resulting from collaborative programs. These strategic alliances balance the Company's exposure to research and development risks inherent in the industry while retaining a share in the success of future products.

A summary of the material terms of agreements that relate to three currently partnered programs follows:

The 1994 Pfizer Agreement – covers the Company’s GABA-based sleep disorder program as partnered with Pfizer

• Pfizer purchased approximately 1.1 million shares of Neurogen common stock for approximately $9.9 million.
• Neurogen received funding from 1994 through 2001 for research under this program. Clinical development of NGD 96-3, a drug candidate discovered under the collaboration, continues under Pfizer management.
• Pfizer has the right to determine when to advance compounds in the clinical process.
• Neurogen will receive milestone payments if specified development and regulatory objectives are achieved.
• Pfizer received the exclusive worldwide license to manufacture, use, and sell GABA-based sleep disorder products discovered during the research collaboration (the period 1994 through 2001).
• Pfizer is required to pay Neurogen royalties based on net sales levels, if any, for such products.

The 2001 Aventis Agreement – covers the Company’s CRF-1-based depression and anxiety program

• Aventis paid $10 million for the licensing of Neurogen’s CRF-1 drug candidates and technology.
• Neurogen receives a total of $11.25 million in committed research payments for the three-year period ending in December 2004.
• Neurogen receives milestone payments if specified development and regulatory objectives are achieved. The company achieved a preclinical milestone in December 2003 and received the preclinical milestone payment of $1 million in January 2004.
• Aventis received the exclusive worldwide license to manufacture, use and sell CRF-1 receptor modulator products developed in the collaboration.
• Aventis is required to pay Neurogen royalties based on net sales of any such products.

The 2003 Merck Agreement – covers the Company’s VR1-based pain management program

• Merck purchased approximately 1.8 million shares of Neurogen common stock for $15 million.
• Merck paid Neurogen a $15 million license fee.
• Neurogen receives $16.8 million in committed research funding and license maintenance payments for the three-year period ending in December 2006.
• Merck has the option to extend the discovery and research effort for up to an additional two years or to terminate the agreement after two years and transfer certain rights to the collaborative program to Neurogen.
• Neurogen will receive milestone payments if specified development and regulatory objectives are achieved. The Company is eligible to receive milestone payments on all VR1 compounds discovered by either company prior to or during the collaboration with the exception of a single early milestone payment on a Merck compound in preclinical development at the time the collaboration was struck. Neurogen is eligible to receive milestone payments on multiple indications.
• Merck received the exclusive worldwide license to manufacture, use, and sell VR1 receptor modulator products developed in the collaboration.
• Merck is required to pay Neurogen royalties based on net sales levels, if any, for such products. Neurogen is eligible to receive royalty payments on all VR1 compounds discovered by either company prior to or during the collaborative research program.

Neurogen incurred research and development costs (net of stock compensation) of $32.8 million, $34.1 million, and $34.5 million for the years ending December 31, 2003, 2002 and 2001, respectively. During those same time periods, the Company recognized research and development revenue from its collaboration agreements (including the agreements described above and other prior agreements), of $4.8 million, $4.3 million, and $3.1 million respectively. Thus, during those time periods, overall research and development revenue, as a percentage of total research and development costs (for both partnered and unpartnered programs), was 15 percent, 13 percent, and 9 percent, respectively.

Patents and Proprietary Technology

Neurogen’s success depends, in part, on the Company’s ability to obtain and enforce patents, maintain trade secrets and operate without infringing the intellectual property rights of third parties. Neurogen files patent applications both in the United States and in foreign countries, as the Company deems appropriate, for protection of both products and processes. As of December 31, 2003, Neurogen is the sole assignee of 222 issued United States patents and numerous foreign patents.

The patent position of biotechnology and pharmaceutical firms is highly uncertain and involves many complex legal and technical issues. There is considerable uncertainty regarding the breadth of claims allowed in such cases and the degree of protection afforded under such patents. As a result, Neurogen cannot assure you that patent applications will be successful or that current or future patents will afford the Company protection against competitors. It is possible that Neurogen patents will be successfully challenged or that patents issued to others may preclude the Company from commercializing its products. Litigation to establish the validity of patents, to defend against infringement claims or to assert infringement claims against others can be lengthy and expensive, even if a favorable result is obtained. Moreover, much of the Company’s expertise and technology cannot be patented.

Neurogen also relies heavily on trade secrets and confidentiality agreements with collaborators, advisors, employees, consultants, vendors and other service providers. The Company cannot assure you that these agreements will not be breached or that trade secrets will not otherwise become known or be independently discovered by competitors. Neurogen’s business would be adversely affected if competitors were able to learn the Company’s secrets or if the Company were unable to protect its intellectual property. Neurogen is not currently party to any contract that would obligate the Company to pay royalties to any third party on any compound in its current portfolio.

Competition

The biopharmaceutical industry is highly competitive and subject to rapid and substantial technological change. Developments by others may render Neurogen’s drug candidates or technologies noncompetitive or obsolete, or we may be unable to keep pace with technological developments or other market factors. Technological competition in the industry from pharmaceutical and biotechnology companies, universities, governmental entities and others diversifying into the field is intense and is expected to increase. Many of these entities have significantly greater research and development capabilities than Neurogen, as well as substantially more marketing, manufacturing, financial and managerial resources. These entities represent significant competition. In addition, acquisitions of, or investments in, competing development-stage pharmaceutical or biotechnology companies by large corporations could increase such competitors' financial, marketing, manufacturing and other resources.

Competitors have developed or are in the process of developing products or technologies that are, or in the future may be, the basis for competitive products. The Company’s competitors may develop products that are safer, more effective or less costly than any Neurogen products or may be able to complete their development more quickly. Neurogen would be at a significant competitive disadvantage if a competitor developed and successfully commercialized a drug similar to a Neurogen drug candidate ahead of the Company’s efforts.

Manufacturing

Neurogen is currently relying, in part, on third-party manufacturers to produce large quantities of development candidate compounds for preclinical development and to produce dosage forms of these candidates to support clinical trials.

Pfizer manufactures or will be responsible for manufacturing, drugs for clinical trials which are subject to the 1994 Pfizer Agreement and has the right to manufacture future products under this collaboration, if any, for commercialization.

Aventis will be responsible for manufacturing, or having manufactured, drugs for clinical trials which are subject to the Aventis Agreement, and has the right to manufacture future products under the collaboration, if any, for commercialization.

Merck will be responsible for manufacturing, or having manufactured, drugs for clinical trials which are subject to the Merck Agreement, and has the right to manufacture future products under the collaboration, if any, for commercialization.

With respect to compounds not currently subject to collaborations, the Company currently utilizes third-party manufacturers for preclinical development and clinical trials.

Sales and Marketing

Neurogen’s strategy is to market any products it develops in the future either directly or through co-promotion arrangements or other licensing arrangements with large pharmaceutical or biotechnology companies. The Company does not expect to establish a direct sales capability for at least the next several years, though it may pursue such a capability in the future. Pfizer has the right to market worldwide future products, if any, resulting from the 1994 Pfizer Agreement. Aventis has the right to market worldwide future products, if any, resulting from the Aventis Agreement. Merck has the right to market worldwide future products, if any, resulting from the Merck Agreement.

Government Regulation

The production and marketing of products Neurogen may develop in the future, as well as research and development activities, are subject to regulation for safety, efficacy and quality by numerous governmental authorities in the United States and other countries. In the United States, drugs are subject to rigorous federal regulation (Food and Drug Administration or FDA) and, to a lesser extent, state regulation. The Federal Food, Drug and Cosmetic Act, as amended, and the regulations promulgated thereunder, and other federal and state statutes and regulations govern, among other things, the testing, manufacture, safety, efficacy, labeling, storage, record keeping, approval, advertising and promotion of our products. Product development and approval within this regulatory framework will take a number of years and involve the expenditure of substantial resources.

The steps required before a pharmaceutical agent may be marketed in the United States include:

1. Preclinical laboratory tests, in vivo preclinical studies and formulation studies,
2. The submission to the FDA of an Investigational New Drug Application (IND) for human clinical testing which must become effective before human clinical trials can commence,
3. Adequate and well-controlled human clinical trials to establish the safety and efficacy of the drug,
4. The submission of a New Drug Application (NDA) or Product License Application to the FDA, and
5. FDA approval of the NDA or Product License Application prior to any commercial sale or shipment of the drug.

In addition to obtaining FDA approval for each product, each domestic drug manufacturing establishment must be registered with, and approved by, the FDA. Domestic manufacturing establishments are subject to biennial inspections by the FDA and must comply with the FDA's Good Manufacturing Practices for both drugs and devices. To supply products for use in the United States, foreign manufacturing establishments must comply with Good Manufacturing Practices and are subject to periodic inspection by the FDA or by regulatory authorities in such countries under reciprocal agreements with the FDA.

Preclinical testing includes laboratory evaluation of product chemistry and formulation, as well as animal studies to assess the potential safety and efficacy of the product. Preclinical safety tests must be conducted by laboratories that comply with FDA regulations regarding Good Laboratory Practices. The results of the preclinical testing are submitted to the FDA as part of an IND and are reviewed by the FDA prior to the commencement of human clinical trials. Unless the FDA objects to an IND, the IND will become effective 30 days following its receipt by the FDA.

Clinical trials involve the administration of the new drug to healthy volunteers or to patients under the supervision of a qualified principal investigator. Clinical trials must be conducted in accordance with Good Clinical Practices under protocols that detail the objectives of the study, the parameters to be used to monitor safety and the efficacy criteria to be evaluated. Each protocol must be submitted to the FDA as part of the IND. Further, each clinical study must be conducted under the auspices of an Independent Institutional Review Board at the institution where the study will be conducted. The Institutional Review Board will consider, among other things, ethical factors, the safety of human subjects, and the possible liability of the institution. Compounds must be formulated according to Good Manufacturing Practices.

Clinical trials are typically conducted in three sequential phases, but the Phases may overlap. In Phase I, the initial introduction of the drug into healthy human subjects, the drug is typically tested for safety (no adverse side effects), absorption, dosage tolerance, metabolism, bio-distribution, excretion and pharmacodynamics (clinical pharmacology). Phase II typically involves studies in a limited patient population to:

1. Determine the efficacy of the drug for specific, targeted indications.
2. Determine dosage tolerance and optimal dosage.
3. Identify possible adverse side effects and safety risks.

When a compound is found to be effective and to have an acceptable safety profile in Phase II evaluations, Phase III trials are undertaken to further evaluate clinical efficacy and to test for safety within an expanded patient population at geographically dispersed clinical study sites. Neurogen or the FDA may suspend clinical trials at any time if it is believed that the individuals participating in such trials are being exposed to unacceptable health risks.

The results of the pharmaceutical development, preclinical studies, and clinical studies are submitted to the FDA in the form of an NDA for approval of the marketing and commercial shipment of the drug. The testing and approval process is likely to require substantial time and effort. The approval process is affected by a number of factors, including the severity of the disease, the availability of alternative treatments and the risks and benefits demonstrated in clinical trials. Consequently, there can be no assurance that any approval will be granted on a timely basis, if at all. The FDA may deny an NDA if applicable regulatory criteria are not satisfied, require additional testing or information or require post-marketing testing and surveillance to monitor the safety of a company’s products if it does not believe the NDA contains adequate evidence of the safety and efficacy of the drug. Notwithstanding the submission of such data, the FDA may ultimately decide that an NDA does not satisfy its regulatory criteria for approval. Moreover, if regulatory approval of a drug is granted, such approval may entail limitations on the indicated uses for which it may be marketed. Finally, product approvals may be withdrawn if compliance with regulatory standards is not maintained or if problems occur following initial marketing.

Among the conditions for NDA approval is the requirement that any prospective manufacturer's quality control and manufacturing procedures conform to Good Manufacturing Practices. In complying with standards set forth in these regulations, manufacturers must continue to expend time, money and effort in the area of production and quality control to ensure full technical compliance. Manufacturing establishments, both foreign and domestic, also are subject to inspections by or under the authority of the FDA and by other federal, state or local agencies.

Whether or not FDA approval has been obtained, approval of a product by regulatory authorities in foreign countries must be obtained prior to the commencement of commercial sales of the product in such countries. The requirements governing the conduct of clinical trials and product approvals vary widely from country to country, and the time required for approval may be longer or shorter than that required for FDA approval. Although there are some procedures for unified filings for certain European countries, in general, each country at this time has its own procedures and requirements.

In addition to regulations enforced by the FDA, Neurogen is also subject to regulation under the Occupational Safety and Health Act, the Environmental Protection Act, the Toxic Substances Control Act, the Resource Conservation and Recovery Act and other present and potential future federal, state or local regulations. The Company’s research and development involves the controlled use of hazardous materials, chemicals, and various low-level radioactive compounds. Although the Company believes that safety procedures for handling and disposing of such materials comply with the standards prescribed by state and federal regulations, the risk of accidental contamination or injury from these materials cannot be completely eliminated. In the event of any accident, Neurogen could be held liable for any damages that result and any such liability could exceed our resources.

Employees

As of December 31, 2003, the Company had 155 full-time employees, of which 123 persons were scientists and, of these scientists, 59 had Ph.D.s or other doctoral degrees. None of the employees are covered by collective bargaining agreements, and the Company considers relations with employees to be good. All current scientific personnel have entered into confidentiality and non-competition agreements with the Company.

Research and Development Expenses

The Company incurred research and development expenses of $32.8 million, $34.1 million and $34.5 million in 2003, 2002, and 2001, which exclude non-cash stock compensation charges of $0.3 million, $0.1 million and $0.9 million, respectively.

Available Information

The Company’s website address is www.neurogen.com. We have included our website address as an inactive textual reference and do not intend it to be an active link to our website. The information that appears on our website is not part of this Form 10-K. Neurogen makes available free of charge through its website all of the Company’s filings with the Securities and Exchange Commission ("SEC"), including its Annual Report on Form 10-K, Quarterly Reports on Form 10-Q, Current Reports on Form 8-K and all amendments to these reports. The Company’s filings are made available as soon as reasonably practicable after such material is electronically filed with the SEC. Copies without exhibits are also available, without charge, from Neurogen Corporation, 35 Northeast Industrial Road, Branford, CT 06405.

ITEM 2. PROPERTIES

The Company conducts its operations in laboratory and administrative facilities on a single site located in Branford, Connecticut. The total facilities under ownership comprise approximately 148,000 square feet, of which 106,000 square feet is in use. Approximately 27,000 square feet has not yet been adapted for the Company’s research and development efforts and approximately 15,000 square feet is currently leased to a third party.

In 1995, the Company leased approximately 24,000 square feet of a 39,000 square foot building under a ten-year agreement, which provided an option to purchase the entire facility effective after the fifth year of the original term of the lease. In January 2001, Neurogen acquired the entire facility for $2.4 million. The additional 15,000 square feet continues to be leased through Neurogen to the third-party tenants who occupied the facilities at the time of purchase.

ITEM 3. LEGAL PROCEEDINGS

The Company knows of no material litigation or proceeding pending or threatened to which Neurogen is, or may become, a party.

ITEM 4. SUBMISSION OF MATTERS TO A VOTE OF SECURITY HOLDERS

None.

CAUTIONARY STATEMENT FOR PURPOSES OF THE
"SAFE HARBOR" PROVISIONS OF THE PRIVATE SECURITIES
LITIGATION REFORM ACT OF 1995

Except for historical matters, the matters discussed in this Form 10-K are "forward-looking statements" within the meaning of the Private Securities Litigation Reform Act of 1995 or any rules, regulations or releases of the Securities and Exchange Commission with respect thereto. Forward-looking statements in this Form 10-K include, but are not limited to, statements in Item 1 under the caption "Neurogen’s Business Strategy" and "Business – Neurogen Drug Programs" with respect to the Company’s various product development programs and statements in Item 7 under the caption "Management’s Discussion and Analysis of Financial Condition and Results of Operations" with respect to the sufficiency of the Company's cash balance to fund planned operations. In addition, the Company may from time to time make forward-looking statements in the future.

Neurogen wishes to caution readers and others to whom forward-looking statements are addressed, that any such forward-looking statements are not guarantees of future performance and that actual results may differ materially from estimates in the forward looking statements. In addition to the important factors described elsewhere in this Form 10-K and the Company’s other filings with the Securities and Exchange Commission, the following important factors, among others, could affect Neurogen's actual future results and could cause such results to differ materially from estimates expressed in any forward-looking statements made by, or on behalf of, Neurogen:

• Failures, difficulties or delays in discovery, research, development, testing, regulatory approval, production and marketing of any of the Company's drug candidates, including without limitation any unanticipated pre-clinical or clinical failures, delays, delays in regulatory approvals, the failure to develop follow-on candidates in a given program, the failure to attract or retain scientific and management personnel, adverse side effects or inadequate therapeutic efficacy or inadequate drug properties which could prevent or slow product development efforts at any stage of product development by delaying or preventing clinical trials, delaying or preventing regulatory approval for commercialization or adversely affecting the risks and potential rewards of individual candidates and the ultimate acceptance by the market. A very significant majority (estimated to be between 80 and 90 percent) of all drugs which enter human clinical trials fail to reach the market.
  

• Vigorous competition within the Company’s anticipated product markets, including without limitation competition from fully integrated pharmaceutical companies, specialty biotechnology companies and platform technology companies, many or all of which may have substantially greater capabilities, experience and resources than the Company.

• Risk that competitors will succeed in developing technologies (including drug discovery techniques) and products that are more effective than those of the Company or that are commercialized prior to similar technologies or products of the Company.
  


• Neurogen’s dependence on decision-making and performance of its corporate partners with respect to research and development funding, preclinical evaluation of drug candidates, human clinical trials of drug candidates, regulatory filings and manufacturing and marketing expertise with respect to most of its most advanced compounds.
  


• Risk that Neurogen’s interests will not coincide with those of its collaborators with respect to the evaluation of candidates, the timing or conduct of clinical development of compounds, the future production of developed products or strategies with respect to development and commercialization of such products.
  


• Risk that actual research and development costs and associated general and administrative costs may exceed budgeted amounts for a variety of reasons, including the uncertainty of product development in the pharmaceutical industry.
  


• Risk that drug targets pursued by the company may prove to be invalid after substantial investments by the Company.
  


• Inability to obtain sufficient funds through future collaborative arrangements, technology transfers, equity or debt financings or other sources to continue the operation of the Company’s business which may require the Company to reduce substantially or eliminate expenditures for product development, to relinquish rights to certain of its technologies or potential products or to cease operations.
  


• Risk that the Company's patents and trade secrets and confidentiality agreements with collaborators, employees, consultants or vendors will be invalidated or not adequately protect the Company's intellectual property.
  


• Uncertainty of the scope and enforceability of patents in the pharmaceutical and biotechnology industries, which purport to enable competitors to restrict others from pursuing certain drug targets.
  


• Risk that the Company may be prohibited or otherwise restricted from working on certain targets relevant to the Company's business.
  


• The Company's dependence upon third parties for the manufacture of its potential products and the Company’s inexperience in manufacturing if the Company establishes internal manufacturing capabilities, each of which could adversely affect the Company’s future profit margins, if any, and its ability to develop and manufacture products on a timely and competitive basis.
  


• Neurogen’s dependence on third parties to market potential products and Neurogen's lack of sales and marketing capabilities, each of which could adversely affect the success of any sales and marketing efforts for the Company's products.
  


• Unavailability or inadequacy of medical insurance or other third-party reimbursement for the cost of purchases of the Company's products.
  


• Inability of the Company to attract and retain qualified management, employees and consultants.

PART II

ITEM 5. MARKET FOR REGISTRANT'S COMMON EQUITY, RELATED STOCKHOLDER MATTERS AND ISSUER PURCHASES OF EQUITY SECURITIES

The common stock of Neurogen is traded on the NASDAQ Stock Market under the symbol NRGN. As of March 1, 2004, there were approximately 202 holders of record of the Company's common stock. No dividends have been paid on the common stock to date, and the Company currently intends to retain any earnings for further development of the Company's business.

The following table sets forth the high and low sales prices for the common stock as reported by NASDAQ.

Equity Compensation Plan Information

The following table sets forth, for the Company's equity compensation plans, the number of outstanding option grants under such plans, the weighted-average exercise price of outstanding options, and the number of shares that remain available for issuance under such plans, as of December 31, 2003.

ITEM 6. SELECTED FINANCIAL DATA

ITEM 7. MANAGEMENT'S DISCUSSION AND ANALYSIS OF FINANCIAL CONDITION AND RESULTS OF OPERATIONS

OVERVIEW

Since its inception in September 1987, Neurogen has been engaged in the discovery and development of drugs. The Company has not derived any revenue from product sales and has incurred, and expects to continue to incur significant losses in most years prior to deriving any such product revenues. Revenues to date have come from six collaborative research agreements, one license agreement and one technology transfer agreement.

Collaborative research agreements have been and are expected to continue to be an important source of funding for the Company. In addition, such arrangements not only drive current revenue (through the recognition of upfront and subsequent license fees, research funding and potential milestone payments over the respective contract periods) but over the longer term also, potential future product revenues in the form of royalties if the agreements result in successful drug development and commercialization. The initialization, expiration and specific terms of such agreements have contributed to, and will continue to cause, significant fluctuations in the Company's recognized revenues and losses.

In December 2003, the Company entered into a collaboration agreement (the "Merck Agreement") with Merck Sharp & Dohme Limited ("Merck") to discover and develop next-generation drugs targeting the vanilloid receptor (VR1) for the treatment of pain. While the impact of the agreement on 2003 results was not significant, it is expected to be a meaningful driver of revenues and liquidity over the next three to five years. The agreement became effective in late December 2003 upon approval by the Federal Trade Commission under the Hart-Scott Rodino Act. In January 2004, Neurogen received $30.0 million from Merck, including $15.0 million in up-front license fees and another $15.0 million for the purchase of 1,783,252 shares of newly issued Neurogen common stock (representing approximately 9% of Neurogen's common shares outstanding). Merck will provide Neurogen with research funding and license payments totaling $16.8 million during the initial three-year term of the contract (subject to Merck's right of termination in the third year). Should Merck exercise its right to extend the collaborative research program for up to two years, Neurogen may also receive further payments of between $2.8 million and $4.2 million per year. In addition, Neurogen will be eligible to receive research, development, and approval milestone payments of up to $118.0 million for the successful commercialization of a collaboration drug for a single therapeutic indication. Beyond that, Merck will pay to Neurogen royalties on the sale of products from the collaboration. Revenue related to this collaboration will be recognized pursuant to the methodology discussed under the "Critical Accounting Policies" section below.

Drug discovery and development activities, excluding stock compensation charges, have accounted for between 84% and 85% of total expenses over the last three years. During 2003, the company incurred significant expenses in conducting Phase IIa exploratory clinical trials in asthma and rheumatoid arthritis for NGD 2000-1, an oral C5a antagonist. In January 2004, at the conclusion of the asthma trial, the Company announced that NGD 2000-1 did not demonstrate a therapeutic benefit for treatment of asthma with respect to the primary predetermined endpoint. The exploratory Phase IIa clinical trial using NGD 2000-1 for the treatment of rheumatoid arthritis is still ongoing.

CRITICAL ACCOUNTING POLICIES

The preparation of Neurogen's financial statements in conformity with generally accepted accounting principles requires management to make certain estimates and assumptions that affect the reported amounts of assets and liabilities and disclosure of contingent liabilities at the date of the financial statements and the reported amounts of revenues and expenses during the reporting period. Management makes estimates in, among others, the areas of revenue recognition, income taxes, stock-based compensation and marketable securities and investments. Estimates are based on historical experience and on various other assumptions that management believes to be reasonable under the circumstances. Actual amounts could differ from those estimates.

The Company believes the following critical accounting policies affect management's more significant judgments and estimates used in the preparation of Neurogen's financial statements:

Revenue Recognition

Each of Neurogen's collaborative research, licensing and technology transfer agreements are significant since the terms of such arrangements may cause the Company's operating results to vary considerably from period to period.

The Company has entered into collaborative research agreements that provide for the funding of specified projects in exchange for the grant of certain rights related to potential discoveries. Revenue under these arrangements typically includes upfront non-refundable fees, ongoing payments for specified levels of staffing for research and milestone payments upon occurrence of certain events. Since the adoption of SEC Staff Accounting Bulletin ("SAB") 101 in 2000, the upfront fees are recognized as revenue ratably over the period of performance under the research agreement. The research funding is recognized as revenue as the related research effort is performed. Revenue derived from the achievement of milestones, each of which represents a substantive stage of development towards a long-term goal such as the nomination of a development or clinical candidate or the start of a specific phase of clinical trials or the filing of a New Drug Application with the Food and Drug Administrat ion, is recognized when the milestone event occurs and collectibility is reasonably certain. Emerging Issues Task Force ("EITF") Issue No. 00-21, "Revenue Arrangements with Multiple Deliverables," became effective for all contracts entered into after June 15, 2003. EITF Issue No. 00-21 addresses, for arrangements with multiple deliverables, how the arrangement consideration should be measured, whether the arrangement should be divided into separate units of accounting and how the arrangement consideration should be allocated among the separate units of accounting. Under EITF Issue No. 00-21, the multiple elements typically included in Neurogen's collaborative arrangements (as discussed above) may not meet the criteria for consideration as separate accounting units and may have to be accounted for as a combined unit. The adoption of EITF Issue No. 00-21 did not have a material effect on the company's results of operations in 2003 since only one collaborative arrangement, the Merck Agreemen t, which became effective upon receiving regulatory approval in late December 2003, was entered into after the pronouncement became effective. Revenue recognized from the Merck Agreement was less than $0.1 million in 2003. The Company will review the Merck Agreement and subsequent collaborative arrangements to determine whether the deliverables included in each arrangement qualify for treatment as separate accounting units.

Research and Development Costs

Research and development costs are expensed when incurred.

Stock-Based Compensation

Neurogen utilizes the intrinsic value method to account for stock-based employee compensation. The Company primarily grants qualified stock options for a fixed number of shares to employees with an exercise price equal to the fair market value of the shares at the date of grant. The Company has also issued restricted stock to key executives and records an expense over the vesting periods. The Company accounts for grants of stock options and restricted stock in accordance with Accounting Principles Board ("APB") Opinion No. 25, "Accounting for Stock Issued to Employees," and, accordingly, recognizes no compensation expense for the options when the option grants have an exercise price equal to the fair market value at the date of grant. The Company has adopted the disclosure-only provisions of Statement of Financial Accounting Standards ("SFAS") No. 123, "Accounting for Stock-Based Compensation" as amended by Statement of Financial Accounting Standards SFAS No. 148, "Accounting for Stock-Based Compensation - - Transition and Disclosure."

The Company occasionally grants stock option awards to consultants. Such grants are accounted for pursuant to EITF Issue No. 96-18, "Accounting for Equity Instruments That Are Issued to Other Than Employees for Acquiring, or in Conjunction with Selling, Goods or Services," and, accordingly, recognizes compensation expense equal to the fair value of such awards and amortizes such expense over the performance period. The fair value of each award is estimated using the Black-Scholes model with certain weighted average assumptions (described in Note 7 to the consolidated financial statements).

Marketable Securities

The Company invests in U.S. government and corporate debt securities. The fair value of these securities is subject to volatility and change. The Company considers its investment portfolio to be available-for-sale securities as defined in SFAS No. 115, "Accounting for Certain Investments in Debt and Equity Securities." Marketable securities at each of December 31, 2003 and December 31, 2002 consisted of debt securities with maturities of one month to approximately five years. Available for sale securities are carried at fair value with the unrealized gains and losses reported as other comprehensive income. Realized gains and losses have been determined by the specific identification method and are included in investment income. Different classifications of the Company's marketable securities pursuant to SFAS No. 115 would possibly result in material impacts to the valuation of the securities and investment income.

Income Taxes

The liability method of SFAS No. 109, "Accounting for Income Taxes," is used to account for income taxes. Deferred tax assets and liabilities are determined based on net operating loss carryforwards, research and development credit carryforwards, and differences between financial reporting and income tax bases of assets and liabilities. Deferred items are measured using the enacted tax rates and laws that are expected to be in effect when the differences reverse. Deferred tax assets may be reduced by a valuation allowance to reflect the uncertainty associated with their ultimate realization. Any subsequently recognized tax benefits relating to the valuation allowance for deferred tax assets would be recorded as an income tax benefit in the Statement of Operations or a charge to Additional Paid-In Capital.

Long-lived Assets

In August 2001, the Financial Accounting Standards Board ("FASB") issued SFAS No. 144, "Accounting for the Impairment or Disposal of Long-Lived Assets." SFAS No. 144 supercedes SFAS No. 121, "Accounting for the Impairment of Long-Lived Assets and for Long-Lived Assets to Be Disposed Of." SFAS No. 144 addresses the financial accounting and reporting for impairment or disposal of long-lived assets. This statement provides that (a) an impairment loss should only be recognized if the carrying amount of a long-lived asset is not recoverable from its undiscounted cash flows, and (b) the measurement of impairment loss should be based on the difference between the carrying amount and the fair value of the asset. It also provides that a long-lived asset (or asset group) should be tested for recoverability whenever events or changes in circumstances indicate that potential impairment has occurred. In addition, it provides for the use of probability-weighted cash flow est imates in the recoverability test. Neurogen adopted the provisions of SFAS No. 144 as of January 1, 2002. The Company will periodically review its long-lived assets for potential impairment and recoverability.

RESULTS OF OPERATIONS

Results of operations may vary from period to period depending on numerous factors, including the timing of income earned under existing or future collaborative research agreements, technology transfer agreements or joint ventures, the progress of the Company's partnered and unpartnered research and development projects, the size of the Company's staff and the level of preclinical and clinical development spending on drug candidates in unpartnered programs. Neurogen believes its research and development costs may increase significantly over the next several years as its drug development programs progress. In addition, general and administrative expenses would be expected to increase to support any expanded research and development activities.

Years Ended December 31, 2003 , 2002 and 2001

The Company's fiscal 2003 operating revenues decreased 57% to $6.8 million from 2002 operating revenues of $15.7 million. The revenue decrease is due primarily to the completion in 2002 of the Company's 1999 Technology Transfer Agreement with Pfizer Inc. ("Pfizer") (described below). License fees decreased from $11.4 million in 2002 to $2.0 million in 2003 due to the recognition in 2002 of the final $9.4 million in license fees earned pursuant to the Pfizer Technology Transfer Agreement. The Company recognized $2.0 million in license fees from Aventis Pharmaceuticals, Inc. ("Aventis") as part of the Aventis Agreement (described below) in both 2003 and 2002. Research revenues increased by 11% to $4.8 million in 2003 from $4.3 million in 2002 due to the recognition of a $1.0 million milestone payment from Aventis upon the successful nomination of a development candidate, offset in part by a $0.5 million decrease in Aventis research funding.

Fiscal 2002 operating revenues increased 37% to $15.7 million from 2001 operating revenues of $11.5 million. The revenue increase consisted of a $3.0 million increase in license fees and a $1.2 million increase in research revenues. The $3.0 million increase in license fees was due to the recognition of a one-time $2.0 million payment from Pfizer to satisfy all remaining obligations related to Pfizer's use of the Accelerated Intelligent Drug Discovery ("AIDD"™) technology under the 1999 Technology Transfer Agreement, a $1.8 million increase in license fees recognized under the Aventis Agreement, and an offsetting $0.8 million decrease in revenue from AIDD™ technology development and improvement services provided to Pfizer. The increase in research revenues was due to a $4.1 million increase in research revenues from Aventis (reflecting a full year of research funding under the companies' December 2001 collaboration agreement) offset in part by a $2.9 million decrease in revenues from Pfizer due to the scheduled conclusion in December 2001 of the companies' joint research collaboration on gamma-aminobutyric acid (GABA)-based drugs.

Research and development expenses, excluding non-cash stock compensation charges, decreased by $1.3 million or 4% to $32.8 million in 2003 from $34.1 million in 2002. Following the implementation of the Company's Operational Excellence program in late 2002 (which was designed to enhance the efficiency of the Company's drug development platform, reduce costs and more closely align the Company's investment of resources with long-term business goals), the Company's research and development related salaries and benefits decreased by $1.4 million (reflecting a 17% reduction in average headcount compared to the prior year) and lab supplies decreased by $2.4 million. These cost reductions were partially offset by increases in expenses related to the Company's exploratory Phase IIa clinical trials (for both asthma and rheumatoid arthritis indications) in its C5a program, chemical manufacturing and other outsourced research studies associated with the preclinical VR1 pain program, as well as patent expenses.

Research and development expenses of $34.1 million, excluding non-cash stock compensation charges, for fiscal 2002 were comparable to $34.5 million in 2001. External development costs related to potential drug candidates decreased from $5.5 million in 2001 to $3.5 million in 2002, due partly to the assumption by Aventis (pursuant to the terms of the Aventis Agreement) commencing in December 2001 of the costs of developing potential drug candidates for treating depression and anxiety disorders. This decrease was offset by an overall increase in the Company's ongoing investment in scientific personnel and research supplies. Research and development expenses represented 85%, 84% and 84% of total operating expenses (excluding non-cash stock compensation charges) for the years ended December 31, 2003, 2002 and 2001, respectively.

The Company expenses all research and development costs as incurred. While the Company maintains a system to record the level of staffing time spent on each of its research and development projects, it does not maintain a historical cost accounting system with sufficient accuracy to reliably estimate its research and development costs on a specific project-by-project basis. A significant portion of the Company's research and development expenses (such as laboratory supplies, travel, information systems and services and facilities costs) benefit multiple projects and are not individually tracked to a specific project. Further, the Company's staff timekeeping system does not account for differences in compensation costs between lower level technicians and more senior scientists.

General and administrative expenses, excluding non-cash stock compensation charges, decreased by 12% to $5.9 million in 2003 as compared to $6.7 million in 2002 and $6.6 million in 2001. The decrease is primarily due to $0.3 million lower legal expenses in 2003 (as the Company incurred higher legal expenses in 2002 in connection with the filing of a shelf registration), $0.1 million lower salaries and benefits (due to an 18% reduction in average headcount compared to the prior year), and $0.2 million lower investor relations expenses, offset by a $0.2 million increase in insurance costs.

Total stock compensation expenses were $1.0 million in 2003 compared to $0.6 million in 2002 and $1.5 million in 2001. Both the 2003 and 2002 stock compensation expenses are primarily composed of continuing non-cash charges for grants of certain stock awards in 1997, 2001 and 2002 to certain officers of the Company. The 2001 expense included non-cash charges to income for grants of certain stock awards in 1997 and 2001, as well as one-time non-cash charges totaling $1.0 million for the modification of certain stock options held by two executive officers. A portion of the stock compensation charges (those relating to the 1997 awards) is variable and as a result may fluctuate significantly in the future.

Other income was $1.0 million in 2003 compared to $1.7 million in 2002 and $4.5 million in 2001. The decrease in 2003 is due to reduced investment income caused by a decline in average cash and marketable securities and lower overall returns on invested funds, partially mitigated by a decrease in interest expense resulting from lower outstanding loan balances. The decrease in 2002 (from 2001) was driven by an increase in interest expense of $1.0 million following a $17.5 million first mortgage debt financing entered into in December 2001 and a decrease in investment income, reflecting both lower average cash and marketable securities balances and lower overall returns on invested funds.

For the years ended December 31, 2003, 2002 and 2001, the Company recognized Connecticut income tax benefits of $0.3 million, $0.3 million and $1.2 million, respectively, in the Statement of Operations. The Company recorded total income tax benefits of $0.3 million, $0.6 million and $2.0 million for the years ended December 31, 2003 , 2002 and 2001, respectively, a portion of which related to employee stock options and was recorded to Additional Paid-In Capital. The benefits are the result of Connecticut legislation which allowed the Company to obtain cash refunds from the State of Connecticut for a portion of research and development tax credits in exchange for foregoing the carryforward of these credits into future tax years.

The Company recognized a net loss of $31.6 million for the year ended December 31, 2003, $23.7 million for the year ended December 31, 2002, and $25.4 million for the year ended December 31, 2001. The increase in the 2003 net loss was primarily due to an $8.9 million decrease in operating revenues. The decrease in the 2002 net loss from the 2001 level was primarily due to the increase in operating revenues.

LIQUIDITY AND CAPITAL RESOURCES

At December 31, 2003 and 2002, cash, cash equivalents and marketable securities in the aggregate were $45.9 million and $79.4 million, respectively. At December 31, 2003, $17.7 million of the marketable securities have maturities greater than one year. However, the Company can and may liquidate such investments prior to maturity to meet its operating needs and strategic and/or investment objectives. The Company's cash and other short-term investment levels decreased progressively throughout 2003 to fund operating expenses as well as $4.8 million in annual