Can industry rescue the National Institutes of Health?

Critique has rained down hard on the National Institutes of Health (NIH) recently.1–5 During a period of increased funding, it was difficult to criticize an institution that was growing rapidly and thriving. Now, many are feeling the pain of a crash diet, which is fraught with guilt, blame, and finger pointing. Legislators are justifying the cuts with criticism, complaining that the public's investments have not resulted in sufficient numbers of medically significant breakthroughs to justify expectations promoted by publicly supported initiatives, including the Decade of the Brain (1990–2000) and the doubling of the NIH budget (1998–2003). As the number of investigators has grown, the funding required to sustain this increasingly large scientific establishment has also expanded. Attendance at the annual meeting of the Society for Neuroscience, for example, grew 20-fold in the past 25 years (to more than 28,000 attendees in 2005). Some investigators are blaming the NIH for failing to meet expectations, arguing that funding has become increasingly politicized, and that earmarks, mandated constraints on what science can be pursued (eg, stem cell policies), and even the clinical research Roadmap Initiative and Clinical and Translational Science Awards represent interference with a scientific meritocracy driven by open competition for RO1 funding. With funding tight, decisions about resource allocation within the NIH are scrutinized mercilessly.

One common refrain has been that the role of industry in biomedical research has not been explored adequately, particularly given that total industry funding of biomedical research in the United States is approximately twice that of the NIH in aggregate. Some state that industry is better positioned to produce treatments for disease, that competition has made them more efficient research machines. Andrew Marks, the Editor-in-Chief of Journal of Clinical Investigation, has even suggested that large clinical studies should be funded by industry and not by the NIH.1 Implied in Marks's statement are three tenets: the primary mission of the NIH is to fund the most exciting basic science, industry would be willing and able to fund the key clinical studies, and industry is more efficient than academia or government in pursuing clinical research. Let us examine each of these points.

First, is the primary mission of the NIH to fund the most exciting, cutting-edge biomedical science regardless of its theme? The stated mission of the NIH is “To uncover new knowledge that will lead to better health for everyone”6; this practical purpose has been the NIH's raison d'être since its creation more than 50 years ago (Table 1). Certainly, basic science is crucial to achieving this goal, but clinical research, and the testing of therapies in particular, will always be required to achieve improvements in health.7 Human lives cannot improve until we have data on humans, and this requires clinical research. Furthermore, future funding is at least partly dependent on there being an obvious return on past investments. Basic science discoveries are essential to identifying new therapies, but the return is usually years away. The NIH is directly responsible to Congress and the President, and ultimately responsible to the American people. If therapeutics are tested only by industry, the institutions that support the NIH will and should give it less credit for advancing health, and this could come at the cost of future funding. Thus, the NIH has a mission that embraces clinical research, and this mission clarifies its utility to its funders.

Table 1. The Purpose of Biomedical Research: From the Original National Institutes of Health Mission Statement

From Topping NH, The United States Public Health Services Clinical Center for Medical Research. JAMA 150:541–545, 1952.

The ultimate purpose is to help provide the practicing physicians of this nation and of the world with better means for ameliorating physical suffering and emotional imbalance, for prolonging human life, and for making all the years of that span more useful both to the individual and to society.

Second, is the agenda of industry aligned with that of the NIH and, more broadly, with the public, such that industry would be motivated to perform the clinical studies of greatest public health import? Industry is primarily motivated to create value (eg, profit) for investors, but profit is often aligned with public health benefit: Treatments are more profitable if they are used more frequently and if people are willing to pay a higher price, which in a free market should be equated with greater value through relief of symptoms or improvement in health. However, key clinical questions do not always involve testing a patentable intervention within an indication that is likely to maximize returns. Among the trials funded by National Institute of Neurological Disorders and Stroke since its inception but before 2000,8 the majority of trials did not test drugs or devices under patent (Table 2). Most treatments tested were already in use, such as endarterectomy for carotid stenosis or warfarin for atrial fibrillation. Furthermore, trials that involved proprietary drugs and devices were often in indications that would have limited financial returns for the patent-holder, or involved the testing of entirely new and risky indications. Industry, in contrast, tends to spend much of its budget on copycat drugs, rather than on truly innovative compounds for new indications. Of the 22 agents approved by the US Food and Drug Administration in the past 5 years for neurological indications,9 17 are essentially copycat drugs (Table 3). More broadly, all of the top 10 best-selling drugs on the market are copycats (Table 4).10 Compared with developing a truly novel drug, the risk for developing a copycat drug is much lower and the financial rewards are just as great and more predictable. Thus, the clinical research agendas of the NIH and industry are distinct. The NIH funds trials that industry would never run, and the public value of the NIH investment in trials is large.8

Table 2. National Institute of Neurological Disorders and Stroke–Sponsored Clinical Trials Completed before 2000, by Primary of Clinical Trials Meeting Inclusion Criteria

Trial Title	Funding Period	Active Intervention	Control Intervention	Target Population	Superior Intervention	Cost of Trial
Trial costs were inflated to 2004 dollars based on the medical care component of the Consumer Price Index. NINDS = National Institute of Neurological Disorders and Stroke.
Proprietary drugs or devices
Brain Resuscitation Clinical Trial I	1979–1983	Thiopental	Placebo	Cardiac arrest survivors	No difference	$3,783,150
National Acute Spinal Cord Injury Study II	1983–1987	Methylprednisolone	Placebo	Acute spinal cord injury	Active	$6,330,642
Deprenyl/Tocopherol in Parkinson Disease	1987–1991	Tocopherol/deprenyl	Placebo	Early Parkinson's disease	Active	$34,015,598
Randomized Study of Nicardipine SAH	1987–1992	High-dose nicardipine	Low-dose nicardipine	Subarachnoid hemorrhage	No difference	$9,793,659
Recombinant Beta Interferon for Multiple Sclerosis	1990–1993	Interferon-β-1a	Placebo	Relapsing-remitting MS	Active	$7,771,364
Valproate for Seizure Prophylaxis after Brain Trauma	1990–1997	Valproate	Phenytoin	Posttraumatic seizures	No difference	$5,475,726
Trial of Org-10172 in Acute Ischemic Stroke	1990–1997	Danaparoid	Placebo	Acute ischemic stroke	No difference	$15,023,629
Tissue Plasminogen Activator in Ischemic Stroke	1990–1999	Tissue plasminogen activator	Placebo	Acute ischemic stroke	Active	$18,774,365
Diazepam for Acute Repetitive Seizures	1992–1995	Diazepam rectal gel	Placebo	Acute repetitive seizure	Active	$1,563,303
Remacemide Inpatient Seizure Evaluation Trial	1992–1995	Remacemide	Placebo	Refractory epilepsy	Active	$1,115,095
Randomized Trial of Tirilazad in Acute Stroke	1993–1994	Tirilazad mesylate	Placebo	Acute ischemic stroke	No Difference	$1,674,068
Total						$105,320,601
Others
Extracranial/Intracranial Arterial Anastomosis Study	1977–1984	Superficial temporal-middle cerebral artery bypass	Medical therapy	Ischemic stroke or transient ischemic attack	Control	$29,198,826
National Acute Spinal Cord Injury Study I	1979–1982	Methylprednisolone, high dose	Methylprednisolone, low dose	Acute spinal cord injury	No difference	$3,105,611
Dilantin for Seizure Prophylaxis after Brain Trauma	1983–1989	Phenytoin	Placebo	Posttraumatic seizures	Active	$2,278,078
Brain Resuscitation Clinical Trial II	1984–1989	Lidoflazine	Placebo	Comatose survivors of arrest	No difference	$5,296,353
Asymptomatic Carotid Artery Stenosis Collaborative Study	1985–1995	Carotid endarterectomy	Medical therapy	Asymptomatic carotid artery stenosis	Active	$43,320,428
Stroke Prevention in Atrial Fibrillation I	1986–1990	Warfarin or aspirin	Placebo	Atrial fibrillation	Active	$16,093,548
North American Symptomatic Carotid Endarterectomy Trial	1987–1997	Carotid endarterectomy	Medical therapy	Symptomatic carotid artery stenosis	Active	$64,033,234
National Acute Spinal Cord Injury Study III	1988–1995	Methylprednisolone, long duration	Methylprednisolone, short duration	Acute spinal cord injury	Active	$12,639,013
Indomethacin Germinal Matrix/Intraventricular Hemorrhage Trial	1989–1993	Indomethacin	Placebo	Very low birth weight neonates	Active	$8,875,272
Brain Resuscitation Clinical Trial III	1990–1993	High-dose epinephrine	Low-dose epinephrine	Cardiopulmonary resuscitation	No difference	$5,753,116
Stroke Prevention in Atrial Fibrillation II	1991–1992	Warfarin	Aspirin	Atrial fibrillation	No difference	$7,941,150
Stroke Prevention in Atrial Fibrillation III	1993–1997	Low-dose warfarin + aspirin	Standard warfarin	Atrial fibrillation	Control	$18,695,305
Pallidotomy in Parkinson's Disease	1993–1998	Unilateral pallidotomy	Medical therapy	Parkinson's disease	Active	$3,007,431
Aspirin and Carotid Endarterectomy	1994–1995	High-dose aspirin	Low-dose aspirin	Carotid endarterectomy	Control	$3,920,504
Nicotine/Haloperidol Therapy in Tourettes	1994–1996	Transdermal nicotine	Placebo	Tourette's disorder	Active	$722,105
Conventional vs Percutaneous Discectomy	1994–1997	Percutaneous discectomy	Conventional discectomy	Lumbar disc herniation	No difference	$3,559,807
Total						$228,439,781

Table 3. Food and Drug Administration–Approved Medications for Neurological Indications (broadly defined) 2001–2005

Brand Name	Generic Name	Indication	Year Approved
Drugs with new mechanism of action for indication
Rozerem	ramelteon	Insomnia	2005
Tysabri	natalizumab	Multiple sclerosis	2004
Neurontin	gabapentin	Postherpetic neuralgia	2002
Strattera	atomoxetine	Attention deficit/hyperactivity disorder	2002
Xyrem	sodium xybate	Narcolepsy	2002
Drugs mimicking an approved drug in the indication
Apokyn	apomorphine	Parkinson's disease	2004
Lunesta	eszopiclone	Insomnia	2004
Lyrica	pregabalin	Neuropathic pain	2004
Cialis	tadalafil	Erectile dysfunction	2003
Levitra	vardenafil	Erectile dysfunction	2003
Namenda	memantine	Alzheimer's disease	2003
Avinza	morphine	Pain	2002
Rebif	interferon-β-1a	Multiple sclerosis	2002
Relpax	eletriptan	Migraine	2002
Axert	almotriptan	Migraine	2001
Bayer Aspirin	aspirin	Migraine	2001
Focalin	dexmethlyphenidate	Attention deficit/hyperactivity disorder	2001
Frova	frovatriptan	Migraine	2001
Metadate CD	methylphenidate	Attention deficit/hyperactivity disorder	2001
Reminyl	galantamine	Alzheimer's disease	2001
Ultracet	tramadol/acetaminophen	Pain	2001
Zomig-ZMT	zolmitriptan	Migraine	2001

Table 4. Ten Drugs with Greatest Worldwide Sales in 2005

Brand Name	Generic Name	Indication	Total Sales (billions)
Based on data from IMS Health for 12 months ending December 2005.10
Drugs with new mechanism of action for indication
None
Drugs mimicking an approved drug in the indication
Lipitor	atorvastatin	High cholesterol	$12.9
Plavix	clopidogrel	Heart disease/stroke	$5.9
Nexium	esomeprazole	Heartburn	$5.7
Advair	fluticasone/salmeterol	Asthma	$5.6
Zocor	vardenafil	High cholesterol	$5.3
Norvasc	amlodipine	High blood pressure	$5.0
Zyprexa	olanzapine	Schizophrenia	$4.7
Risperdal	risperidone	Schizophrenia	$4.0
Prevacid	lansoprazole	Heartburn	$4.0
Effexor	venlafaxine	Depression	$3.8

Finally, is industry-sponsored research more efficient than that funded by the NIH? A key conservative tenet is that competition drives efficiency and productivity, and that governmental organizations suffer from lack of competition. There are no data that we can find that answer this question. There has been substantial growth in NIH funding in the past 10 years without a commensurate increase in Food and Drug Administration–approved therapies, but industry investment in research has increased even more dramatically during this period.11 We currently have no real metrics for efficiency of clinical trials or, even more daunting, of translational or basic research. Without these, all sectors are open to criticism.

Of course, the position that the NIH should not fund major clinical research projects is rather extreme and easy to dispute. It is much harder to find the appropriate balance between clinical and basic funding, and the most rational boundary between what is funded by industry and what is the responsibility of the NIH. The frontier between NIH and industry will certainly change over time, particularly if private-sector funding continues to grow at its current rate and public sector funding remains stagnant. Perhaps more important than defining a clearly demarcated perimeter is understanding how to move things across it.

There is no argument here that things could be improved. Better cooperation between the NIH and industry could make both more efficient and productive and certainly would be in the public's best interest. Most of the trials based on NIH-funded discoveries will be sponsored by the biomedical industry, and most of the profit will be theirs also. Health insurers and the businesses paying their premiums also have something to gain, and sometimes lose, from the biomedical research investment. All these institutions could benefit from reexamining research priorities, identifying common barriers to productivity, and setting strategies to overcome these barriers through directed funding or policy proposals. The Roadmap Initiative currently consumes 1% of the total budget of the NIH. The Roadmap's focus on a narrow definition of translational research—from identification of disease targets to proof of principle at the bedside—is an ideal linchpin for the creation of a new and larger public–private initiative. Matching contributions from the pharmaceutical and health care industries, combined with Roadmap dollars, could be used to create a new engine for growth of translational science, and for cooperation, at a time of historic belt-tightening by all. Not a new idea, but perhaps one whose time has come.