RESEARCH and development of new anesthetic drugs by the pharmaceutical industry has greatly improved patient care and transformed clinical practice. Sharing the spotlight with industry has been the Food and Drug Administration (FDA), which assures these new agents are safe and effective. In this article, we discuss how the FDA approves and licenses investigational drugs.
The Anesthetic & Critical Care Drugs section of the FDA is responsible for all local and volatile inhalational anesthetics, potent sedatives/hypnotics, muscle relaxants, and opioid analgesics. In addition to medical officers (anesthesiologists, anesthesiologist‐intensivists), the team consists of pharmacologists, pharmacists, pharmacokineticists, chemists, statisticians, epidemiologists, and product managers (administrative coordinators).
Food, Drug, and Cosmetic Act
Enacted in 1938 and expanded in 1962, the Food, Drug, and Cosmetic Act prohibits interstate shipment (i.e., administration to human subjects) of unapproved drugs unless the manufacturer receives an exemption from the FDA under the terms of a Notice of Claimed Investigational Exemption for a New Drug (IND). After this has been granted, the manufacturer can commence clinical trials “under the IND.” Permission to market an investigational drug is granted only after data obtained from these trials meet the safety and efficacy criteria required for New Drug Application (NDA) approval.
Although most INDs are filed by pharmaceutical or biotechnology companies (or sponsors, in regulatory terminology), individual investigators may do so as well. In general, the purpose of these noncommercial protocols is to study an indication, a patient population, a route of administration, and/or a dose not described in the label (package insert) of a marketed drug. For example, a recent investigator‐IND questioned whether there were gender differences in the analgesic effectiveness of escalating doses of intrathecal neostigmine. Although individuals who receive an IND must conform to the record keeping and reporting requirements mandated by the FDA, not all protocols require an IND, provided certain criteria are met (Table 1
During the early phases of an IND, what matters most to the FDA is ensuring the safety of human subjects. In later stages, robust trial design (hypothesis testing, adequate sample size, appropriate controls, patient selection and assignment, control of bias, and statistical analysis) assumes equal importance (Table 2
) and constitutes the basis for establishing efficacy (Table 3
). For a given NDA, therefore, the final regulatory decision‐"approved," “approvable, provided certain conditions are fulfilled,” or “nonapprovable”‐is largely a matter of how confident the agency is that the drug is safe and effective for the proposed indications.
Drug manufacturers typically request two major conferences with the Anesthetic and Critical Care Drugs section. The first is the “pre‐IND” meeting, in which the sponsor presents preclinical (animal) data and outlines the clinical trials it proposes to conduct. Based on the agency's comments and suggestions, the sponsor submits its IND application (form 1572). Form 1572 must supply information to guide it to the appropriate section within the agency, provide adequate pharmacologic and toxicologic animal data to enable a safety assessment to be made, and submit a method of recalling the drug should the investigation reveal a risk to human subjects. Unless the applicant receives notice to delay a proposed investigation because the agency believes it poses an unacceptable risk to human subjects (an order known as a clinical hold), the IND application is approved and clinical trials may begin.
As the clinical portion of the drug development plan nears completion, a “pre‐NDA” conference is scheduled. At this meeting, the sponsor presents data from IND trials that it believes demonstrate the product's effectiveness (termed pivotal trials) and solicits the agency's comments regarding any additional information that may be required before it formally submits the NDA “package” for consideration.
Questions also can be submitted to the agency in writing or during a telephone conference at any time during this entire process.
For clinical trials conducted under an IND, the sponsor must notify the agency by telephone within 7 calendar days of any serious adverse drug event (SAE), defined as fatal, life threatening, permanently disabling, resulting in congenital anomaly, or requiring/prolonging medical intervention and inpatient hospitalization. In addition, a written IND safety report must be submitted to the FDA and all participating investigators with 15 calendar days. The FDA expects sponsors to submit written IND safety reports each time the sponsor receives information about an SAE whether it is described in the investigator's brochure or not ("unexpected" SAE).
Safety issues arising during clinical trials can influence drug development, sometimes in a critical way. The local anesthetic sameridine (Astra AB, Sodertalje, Sweden) is a recent case in point. After allocating substantial time and research capital in organizing multicenter clinical trials, the manufacturer elected to discontinue the project because of infrequent, but clinically significant, episodes of sinus bradycardia associated with intrathecal drug administration.
IND Clinical Trial Phases
Federal regulations detail a clinical trial hierarchy. These stages represent a progression that broadens as it goes on, both in the type of studies pursued and in the size and heterogeneity of the population tested. 
1. Phase I trials determine whether a drug is safe to administer to humans. These are typically open‐label, escalating dose‐response pilot studies conducted in up to 25‐50 healthy volunteers or patients. Although phase I protocols need only provide a general outline of the proposed investigation, elements critical for safety, e.g., monitoring, blood chemistries, and toxicity‐based stopping or dose adjustment rules, must be described in detail.
2. Phase IIa trials seek to confirm that a drug is effective for the purpose intended. They comprise up to 100 ‐ 200 patients and are either open‐labeled or blinded. After their completion, sponsors often will convene a data monitoring board of outside consultants to review the (blinded) data. If there are serious efficacy or safety concerns, the drug may be withdrawn from ongoing development.
Agents that continue to hold promise are entered into phase IIb investigations. Their purpose is to identify active metabolites, optimal effective dose, optimal dose interval, and other pharmacokinetic properties of the drug.
3. Phase III trials assess drug safety and efficacy in large patient populations (several thousand). To receive NDA approval, data from at least two adequately designed and well‐controlled pivotal trials must show effectiveness for the proposed indications. Federal regulations require that racial minorities and three subpopulations‐women, children, and elderly persons‐be sufficiently represented.
4. Phase IV trials are undertaken after NDA approval. The agency requests them whenever a new drug appears to be safe and effective in the general population but insufficient data exist in a specific subpopulation, e.g., neonates. Stipulations for their conduct are discussed with the sponsor before granting approval and are outlined in the official approval letter (phase IV “commitments”). Sponsors also launch phase IV trials on their own to obtain additional data for future marketing purposes, such as a new indication.
The NDA “Package”
During the past 5 years, a number of anesthesia‐related drug products have received NDA approval (Table 4
). The submission for remifentanil (Ultiva, Glaxo Wellcome, Research Triangle Park, NC) illustrates the scope of IND clinical trials required to market an investigational drug. [Section]
Remifentanil was approved by the FDA in July 1996. This decision was based on extensive review of data from approximately 3,000 patients and 300 volunteers, in which the drug was studied as a component of general anesthesia, for use in monitored anesthesia care, and for short‐term postoperative analgesia in the postanesthetic recovery unit. It was used in combination with thiopental and propofol for induction of general anesthesia, in combination with nitrous oxide, propofol, and isoflurane for maintenance of general anesthesia, and alone and in combination with midazolam for monitored anesthesia care. Remifentanil generally was administered in higher dosages (ED90 vs. ED50) than the comparator opioid‐alfentanil in most trials, fentanyl in the remainder; although physiologic responses to noxious stimuli were lower in remifentanil subjects, recovery times for the two groups were similar. In terms of its safety profile, recovery of respiratory function generally was rapid; however, rare cases of respiratory depression occurred as long as 25 min after discontinuation of remifentanil when used in combination with isoflurane. In addition, several cases of respiratory depression and apnea occurred when remifentanil was infused inadvertently, e.g., with flushing of the intravenous line during administration of another drug or from residual remifentanil left in the tubing after discontinuation. For these reasons, NDA approval was granted with the stipulation that the label clearly indicate that administration is suitable only in highly monitored settings (operating room or postoperative recovery room) under the supervision of the anesthesia practitioner. In addition, because of the extremely rapid onset and termination of action, the sponsor was charged with launching an educational program that emphasized safety issues in the infusion of remifentanil, before supplying it to practitioners. Every 3 months, the sponsor submitted a report updating the progress of this program.
Using the MedWatch reporting system, the FDA must be notified on a regular basis of all adverse events (AEs) associated with drug administration. 
(As of September 1998, MedWatch reports can be transmitted directly to the FDA via the internet: https://http://www.accessdata.fda.gov
/medwatch/medwatch‐online.htm.) Notice of serious adverse events received by the sponsor from practitioners or hospitals must be forwarded to the agency within 15 calendar days of receipt. These regulations remain in effect as long as the drug is marketed in this country. (See reference 2 for a more extensive discussion of the MedWatch system.)
Each MedWatch report is entered into a computer database and screened by a member of the agency's Division of Epidemiology and Pharmacovigilance. When a suspicious pattern emerges, the reports are submitted to a medical officer, who may elect to telephone the MedWatch originator and obtain a detailed history. If these suspicions are confirmed, the evidence and, when necessary, the need for a labeling change is discussed with the sponsor. This is often accompanied by a request that the sponsor distribute a “Dear Doctor” letter to physicians, hospital pharmacists, and other healthcare professionals, informing them of the labeling change. Depending on the circumstances, the agency can undertake still sterner actions: initiate a recall (temporary removal of a marketed product) or, rarely, withdraw approval of the drug in question.
Paraplegia associated with low‐molecular‐weight heparin administration in patients undergoing continuous epidural anesthesia is one of the most disturbing serious adverse events ever reported to MedWatch. It is worth noting that these agents were first approved in 1993 for deep venous thrombosis prophylaxis by medical officers in the Division of Anticoagulant and Gastrointestinal Drugs, who were unfamiliar with the techniques of regional anesthesia. During the next 4 yr, as the number of drug exposures increased and newer low‐molecular‐weight heparins appeared on the market, MedWatch received more than 30 reports of anesthesia‐related neuraxial hematomas. In the fall of 1997, the Anesthesia and Critical Care Drugs section was consulted to evaluate these reports and provide a clinical context. The FDA subsequently (November 1997) issued a Dear Doctor letter to anesthesiologists, surgeons, and other members of the healthcare community that summarized these cases. A copy of this letter was faxed to the editors‐in‐chief of Anesthesiology and Anesthesia and Analgesia, who published it in their respective journals. In February 1998, a meeting of the Anesthesia and Life Support Advisory Committee (see Links with the Academic Community, below) was held with representatives of several low‐molecular‐weight heparin manufacturers, which resulted in major labeling changes for these agents.
Recommended clinical guidelines for neuraxial anesthesia in the setting of low‐molecular‐weight heparin administration emerging from an American Society of Regional Anesthesia consensus meeting were reported in a supplement to the November/December 1998 issue of Regional Anesthesia and Pain Medicine.
Part of the FDA's mission is to ensure that prescription drug information provided by drug firms is truthful, balanced, and accurately communicated. Federal regulations require that applicants submit specimens of all journal advertisements, scripts for audiovisual testimonials, presentations at medical meetings, advertising literature used by sales representatives, and other drug‐related promotional material to the FDA before initial dissemination. In a small percentage of cases, advertising claims ("safer than") can be misleading when compared to information contained in the label. After joint discussion with the sponsor and the FDA's Division of Drug Marketing, Advertising, and Communication, the applicant may be asked to modify one or more of these items. Although not required, pharmaceutical companies usually comply with these recommendations, rather than risk legal action by the agency.
Other Aspects of FDA Policy
Links to the Academic Community
One way the FDA maintains contact with the academic community is through the Anesthesia & Life Support Advisory Committee. Committee members are selected from leading academic centers and receive term appointments for 3‐5 yr.
Advisory Committees are funded by, but distinct from, the FDA. Operating under specific regulations protecting their autonomy, they advise the agency about matters regarding drug products under development, new drug approvals, and safety concerns relating to products already on the market. Except when proprietary information is being discussed, meetings are open to the public. A financial disclosure statement, updated before each meeting, is required for all committee members. Depending on the degree of perceived conflict‐of‐interest, if any, members can be prohibited from participating in the discussion or casting a vote, or both.
For most NDAs, the majority vote among committee members is consistent with subsequent regulatory action by the agency. Rarely, does the opposite situation arise, e.g., oral transmucosal fentanyl citrate (Actiq; Anesta Corp., Salt Lake City, UT) indicated for breakthrough cancer pain. Although the Advisory Committee's unanimous recommendation was “approvable, as long as the sponsor provides an adequate risk management program to avoid accidental ingestion of the product by children”; the agency's subsequent decision was “nonapprovable.” It is likely that the agency will reconsider this decision after additional efficacy and safety data are submitted by Anesta. On November 5, 1998, Actiq was approved by the FDA.
Informed Consent and Investigational Review Boards
At the Nuremberg Doctors' Trial (1946‐1947), the world learned of pseudoscientific medical experiments conducted on slave laborers by some of Germany's leading physicians. Ironically, informed consent for experimental research on human subjects had been a requirement of the German Ministry of Health 
since 1900, whereas it did not achieve broad support in the U.S. until disclosure in 1972 of the infamous “Tuskegee Study of Untreated Syphilis in the Negro Male.” 
Two years later, the federal government enacted a set of comprehensive rules designed to protect subjects who participate in government‐funded research. 
Informed consent was one cornerstone of the new rules; mandatory investigational review boards (IRB) was the other. As a result, INDs must provide written informed consent for participants and receive IRB approval. The FDA has authority to inspect IRBs, and review and copy IRB records to determine whether an IRB is operating in accordance with its own written procedures and in compliance with current FDA regulations affecting IRBs.
Although there are FDA regulations for informed consent (Table 5
), it is the responsibility of each IRB to ensure that clinical trials protect the rights and well‐being of research subjects. IRBs also are responsible for reviewing how subjects are recruited. For instance, a sponsor may not make claims for a “medical breakthrough” or “free medical treatment.”
Although INDs do not have to be conducted in the U.S., they must meet guidelines enumerated in the Declaration of Helsinki or an equivalent standard of informed consent. If a foreign trial is to be relied on as providing substantial evidence of safety or effectiveness, or both, it must be conducted in a fashion that substantiates the validity of the claim, and the primary records must be available for audit by the agency's Division of Scientific Investigations.
Labeling Changes (Supplemental NDAs)
The drug label serves two important functions: it presents the indications for which a drug is approved, and it summarizes safety and efficacy information obtained from clinical trials conducted by the sponsor (Table 6
). Claims or indications not found in the label have not been reviewed by the FDA and may not meet the agency's standards for safety and efficacy.
Labeling changes sought by a sponsor for a drug already on the market are known as supplemental NDAs. Evidence to support the proposed modification (new route of administration, population, indication, dose) can come from one or more sources. For drugs that have a well‐established “track‐record,” sponsors simply may be asked to provide the results of a literature search of clinical trials published in peer‐reviewed journals that demonstrate safety and effectiveness. This approach was used in 1998 for rectal administration of methohexital (Brevital, Eli Lilly, Indianapolis, IN) in pediatric patients, formerly approved only for intravenous and intramuscular injection. In contrast, sponsors of newer products may be required to conduct formal clinical trials so the agency can review the findings on a line‐by‐line, patient‐by‐patient basis. Before the manufacturer of midazolam (Versed, Hoffman‐LaRoche, Nutley, NJ) received approval in 1997 for a new indication, continuous intravenous administration for sedation of mechanically ventilated adult intensive care unit patients, it had to conduct 3 clinical trials and submit the results to the agency for review.
The agency also may request labeling changes when new safety issues are reported to MedWatch or in the medical literature, or both. In 1996 the agency received a string of MedWatch reports describing acute pancreatitis of unknown cause presenting < 12 h after propofol administration (Diprivan, Zeneca Pharmaceuticals, Wilmington, DE) in otherwise healthy ASA I outpatients undergoing brief, peripheral procedures, such as hardware removal. Each person who submitted a report was contacted by telephone by one of the authors (L.L.), and the narrative was verified for accuracy. Although causality could not be proved, only suspected, a labeling change in the WARNINGS section of the label listing this complication was warranted.
Exclusivity Determination and Generic Competition (Abbreviated NDAs)
Approved drugs receive the right of exclusivity. Exclusivity is an award closely resembling patent protection that allows the promotion and marketing of a given drug for 5 yr from the date of approval. In special situations, the manufacturer may be granted an extension of exclusivity, as defined by congress in the Waxman‐Hatch Act of 1984. For example, a short time after the propofol NDA was approved in 1989, outbreaks of septic shock in previously healthy patients who had received the drug were reported to the agency. The etiology of this epidemic was ultimately traced back to inappropriate handling, i.e., bacterial contamination. To improve propofol's safety profile, the agency recommended (in addition to an educational program for the medical community) that clinical trials be conducted comparing the safety and efficacy of the existing formulation with a sponsor‐proposed formulation containing the bacteriostatic agent EDTA. In return for completing these trials, the FDA granted Zeneca a 3‐yr extension of exclusivity.
After the exclusivity period expires, a drug may be marketed by competitors for the same indications under the generic drug name or an alternative trade name, but not under the originator's trade name. To market a generic anesthetic drug, an “abbreviated NDA” (ANDA) must be filed. For abbreviated NDAs, safety and effectiveness are assumed if the new drug is shown to be bioequivalent (as defined by federal regulations) to the reference drug. The active ingredient must be the same, whereas inactive ingredients may be the same or similar. Competitors generally apply to the originating manufacturer for a “right of reference” to the data from the original pharmacology and preclinical studies to avoid having to repeat them for approval of the generic formulation. When an application for approval of a generic competitor is filed, the Office of Generic Drugs reviews all submitted documentation and makes a side‐by‐side comparison between the proposed generic label and the originator's label to verify that the indication claims are identical and that any formulation or manufacturing changes have been justified. For anesthetic drugs, the Anesthetic and Critical Care section also reviews the generic application and provides a consult to the Office of Generic Drugs as part of the approval process.
Using a drug for an indication, a patient population, a route of administration, or a dose that is not described in the label is considered unapproved or “off‐label” use. Unapproved does not mean disapproved‐the FDA regulates drug products, not the practice of medicine‐and off‐label use of a drug is not necessarily considered medical experimentation (Table 7
The anesthesia community generally learns about “off‐label” uses of approved agents from reports in the literature. Even if these techniques are widely used and come to be considered the standard of practice, e.g., isobaric intrathecal bupivacaine, there may be little financial incentive for the sponsor to file a labeling change for a new indication if the drug is already doing well in the marketplace or has lost exclusivity.
Studies in Women
Since 1993, the FDA has promoted greater inclusion of women in clinical trials to identify possible gender‐related differences in pharmacokinetics and drug‐related responses. 
It is the FDA's position that women of child‐bearing age can safely participate in phase I and early phase II trials before completion of all animal reproduction studies, so long as the informed consent discloses the extent to which preclinical reproductive toxicology and teratogenicity data are known. A negative pregnancy test result and use of an effective method of contraception in premenopausal subjects are required in nearly all anesthetic drug protocols.
Pediatric Population Studies
The pediatric population, defined as birth to 16 yr of age, represents the largest patient subset formerly excluded from the drug investigation process. Approximately 80% of drugs listed in the Physician's Desk Reference contain disclaimers regarding lack of data in children or restricting use to certain age groups. Even “newer” agents, such as fentanyl and bupivacaine, were not investigated in young children as part of the approval process until unexpected toxicities were observed. [7,8,9]
Inadequate pediatric labeling thus exposes children to the risk of unexpected adverse reactions or lack of optimal treatment.
The agency first addressed this problem 
in 1994. Under the “Pediatric Labeling Rule,” if a drug manufacturer determined that existing data permitted modification of the label's pediatric use section, it had to submit a supplemental NDA seeking approval of the labeling change. As long as the course of the disease and the effects of the drug were sufficiently similar in children and adults to permit extrapolation from adult effectiveness data, controlled clinical trials in the pediatric population were not required.
The FDA Modernization Act of 1997 substantially changed this policy. In exchange for conducting new safety and efficacy trials in the pediatric population, sponsors of drugs already on the market receive an additional 6 months of exclusivity. The FDA has published a draft list of those drugs that would most benefit from this type of data. (For additional updates on this topic, see http://http://www.fda.gov
/cder/guidance/pedrule.htm and http://http://www.fda.gov
/cder/pediatric/wrlist.htm.) Sponsors of new NDAs are required to conduct pediatric trials as part of their drug developmental plan, for which they, too, receive an additional 6 months of exclusivity.
Recent Regulatory Changes
The Prescription Drug Users Fee Act, approved by Congress in 1992 and renewed as part of the FDA Modernization Act of 1997, allows the agency to charge drug manufacturers a filing fee (approximately $250,000) to help defray the costs of an NDA review. In return, pharmaceutical companies receive assurance of a regulatory decision within 6‐12 months. Prescription Drug Users Fee Act fees apply only to NDAs‐there is no fee for an IND review‐and cannot be used by FDA to fund ancillary activities related to the drug product under review, e.g., postmarketing safety surveillance.
Former acting FDA Commissioner Friedman recently stated, “Our workload, in terms of new product applications, is increasing by 12% per year. This means it doubles every 6 years.” 
Most, if not all, of the $187 million increase in FDA's fiscal year 1999 budget proposal is earmarked for food safety and tobacco prevention initiatives. Effectively there are no additional funds for fiscal 1999, making the agency increasingly dependent on industry‐derived Prescription Drug Users Fee Act fees to do its job. [12,13]
This predicament in which the FDA now finds itself makes it particularly vulnerable to special interest groups, whose priorities are not necessarily consistent with those of the agency.
Many healthcare policy academics, patient advocate organizations, and consumer “watchdog” groups are alarmed by this situation. They charge that new regulatory policies (paragraphs 1‐4 below) incorporated into the FDA Modernization Act of 1997 signal a change in FDA's role. Rather than regulating industry to protect the health of consumers, the agency has become industry's partner.
1. Information authorized to be disseminated (Sec 551a, 552a): For the first time, manufacturers may promote unapproved ("off‐label") use of their products “directly to a health care practitioners, pharmacy benefit manager, health insurance issuer, (or) a group health plan …” In other words, sponsors may make claims based solely on articles in the medical literature (that typically include far too few subjects to meed FDA criteria for safety and efficacy), provided a disclaimer appears on the promotional material, which states that “the information concerns a use of a drug or device that has not been approved or cleared by the FDA.”
2. Scientific advisory panels (Sec 120n): For the first time, a “representative of … the drug manufacturing industry not directly affected by the matter to be brought before the panel” will be a voting member of each advisory committee.
3. Number of (pivotal) clinical investigations required for approval (Sec 115a): The approval requirement has been reduced from “at least 2 adequate and well‐controlled clinical investigations,” the standard for nearly all drugs approved in the past, to “at least 1 adequate and well‐controlled clinical investigation.”
4. “Accredited third‐party reviewers” (Sec 523b): For the first time, a device manufacturer may submit an application either to the FDA or to an FDA‐accredited “third‐party reviewer … who is an independent organization … not owned or controlled by (the) manufacturer,” selected by the manufacturer seeking approval. Since compensation is paid “by the person who engages such services,” and because the manufacturer can hire these same private consultants to perform additional work in the future, ensuring the “independence” of third‐party reviewers is problematic, at best.
Drug regulation, similar to the practice of medicine itself, is a complex process based on science, coupled with sound clinical judgment and a sense of proportion. This year, the FDA will be asked to approve more new drugs than ever before. As the agency's budget makes clear, there is little increased funding for this effort. It must be accomplished with existing resources.
For decades, the FDA has struggled to protect Americans from the health risks of drugs. By almost any measure, the net result of regulatory changes recently enacted by congress is a reduction in FDA's authority over marketing and development of potentially hazardous products. To what extent the agency can emerge from this latest challenge with its effectiveness‐and integrity‐intact, remains to be seen.
The authors thank Robert A. Bedford MD and Beverly K. Phillip MD for manuscript review.
[Section] Palmisano B, Landow L: From the FDA: Ultiva and Naropin. Anesthesiology 1997; 86:33A‐4A
© 1999 American Society of Anesthesiologists, Inc.