Regulatory Aspects of PET Radiopharmaceutical Production in the United States

Abstract

The Food and Drug Administration (FDA) Modernization Act (the Modernization Act) of 1997 directed FDA to establish current good manufacturing practice (CGMP) requirements for positron emission tomography (PET) drugs. As directed by the US Congress in the Modernization Act, the costly CGMP regulations which the FDA applies to large pharmaceutical manufacturers are not appropriate for the PET drugs due to the unique properties of these drugs mainly the short half-lives. In consideration of the unique nature of PET drugs and PET drug production, FDA instituted specific CGMP requirements in 21 Code of Federal Regulations (CFR) part 212. Sections 1 and 2 of this book chapter illustrate the significant different aspects and the rationales for such differences between part 212 and parts 210/211 which are the CGMP requirements for non-PET drugs. The PET drug CGMP regulation found in part 212 also provides a more flexible regulatory framework for investigational PET drugs for human use produced under an investigational new drug application (IND) in accordance with part 312 and PET drugs produced with the approval of a Radioactive Drug Research Committee (RDRC) in accordance with part 361. The PET CGMP requirements for these PET drugs can be met either by compliance with part 212 or by producing such drugs in accordance with the 32nd edition of the United States Pharmacopeia (USP) General Chapter <823>, “Radiopharmaceuticals for Positron Emission Tomography – Compounding,” which was published in 2009. In 2012, USP revised and renamed this General Chapter <823>, “Positron Emission Tomography Drugs for Compounding, Investigational, and Research Uses.” FDA is currently considering whether to amend the PET CGMP regulations to incorporate this revised chapter into part 212. Section 4 describes format/content of the revised USP General Chapter <823> which is very much in line with part 212 and offers more flexible requirements than the <823> in 32nd USP.

FDA expects to perform a CGMP surveillance inspection of each PET facility once every 2 years on average, but may visit some sites more than once every 2 years when warranted, such as when the site is named in an application for a new type of PET drug product or has undergone substantial change. As part of the drug approval process, a preapproval inspection (PAI) will be performed by the FDA inspector(s) to provide assurance that a PET drug production facility that is named in a drug application is capable of producing the PET drug in accordance with CGMPs and that the submitted application data are reliable, accurate, and complete. Section 5 summarizes the major deficiencies identified by the FDA inspectors during their 2013 and 2014 inspection programs at various PET drug production facilities in the United States.

1 Introduction

Positron emission tomography (PET) radiopharmaceuticals are unique among radiopharmaceuticals because of their short half-lives. There are four positron-emitting radioisotopes that are considered the biologic tracers, carbon-11, nitrogen-13, oxygen-15, and fluorine-18. ¹¹C (t _½ = 20.4 min), ¹⁵O (t _½ = 2.1 min), and ¹³N (t _½ = 10 min) are referred to as the essentials of life. They can be easily substituted directly into biomolecules without changing the properties of the molecule. ¹⁸F (t _½ = 110 min) is not a normal constituent of biological molecules but can often be substituted for a hydroxyl group as in the case of deoxyglucose or can be substituted for a hydrogen atom in a molecule or placed in a position where its presence does not significantly alter the biological behavior of the molecule. Fluorodeoxyglucose (¹⁸FDG) is currently the most widely used PET radiopharmaceutical in clinical oncology in addition to its clinical applications in cardiology and neurology. ¹⁸FDG is one of the important factors for the wide acceptance of PET imaging in clinical practice. The application of PET in clinical oncology is increasing since many molecular targets relevant to cancer can be labeled with positron emitter radionuclides.

The concept of PET and the design and construction of PET scanners were initiated in the United States back in the late 1950 [1–3]. However, the benefits of PET imaging were not widely available to American patients for many years due to “lack of reimbursement and inappropriate and costly regulations promulgated by FDA” [4]. The same current good manufacturing practice (CGMP) requirements that FDA typically applies to large drug manufacturers were also placed on the production of PET radiopharmaceuticals without taking into consideration the special properties of these drugs (e.g., short half-lives). In order to make PET imaging technology more accessible to the US patients at reasonable cost, the US Congress establishes a new regulatory framework for PET drugs within the Food and Drug Administration Modernization Act of 1997 (FDAMA) which was signed by President Bill Clinton on November 21, 1997 [5].

Specifically, section 121(c)(1)(A) of 1997 FDAMA directed FDA to establish CGMP requirements for PET drugs [5]. In the Federal Register (FR) of September 20, 2005 (i.e., 70 FR 55038), FDA published a proposed rule to establish CGMP requirement for PET drugs [6]. On December 10, 2009, FDA issued a final rule in 74 FR 65409 [7] and an accompanying guidance document to establish specifically “tailored” CGMP requirements for PET drugs [8].

These requirements are codified in the Title 21 of the Code of Federal Registrations (CFR) part 212 (part 212) [7]. The deadline for this regulation to become effective was originally set on December 12, 2011 [7]. Two years later, with the deadline looming, FDA received numerous requests for extending the submission deadline. FDA was concerned about preventing access to the PET drugs, and therefore this deadline was extended by the FDA to June 12, 2012 [9].

2 PET Drug

2.1 Is PET Drug for Imaging or Therapy or Both?

“PET drug” as defined in part 212 is “a radioactive drug that exhibits spontaneous disintegration of unstable nuclei by the emission of positrons and is used for providing dual photon positron emission tomographic diagnostic images” (emphasis added) [10]. Under section 121(a) of the FDAMA (codified as section 201(ii)(1)(A) of the Food, Drug, and Cosmetic Act [FD&C]) [5, 11], a compounded PET drug is defined as a drug that “exhibits spontaneous disintegration of unstable nuclei by the emission of positrons and is used for the purpose of providing dual photon emission tomographic diagnostic image” (emphasis added). The word “diagnostic” in each of these two definitions appears to indicate that part 212 [10] and the provisions of the FDAMA concerning PET drugs [5] do not apply to PET drugs used for therapeutic purposes.

2.2 Definition of PET Drug

The definition of PET drug as stated in part 212 specifically includes any nonradioactive reagent, reagent kit, ingredient, nuclide generator, accelerator, target material, electronic synthesizer, or other apparatus or computer program to be used in the preparation of a PET drug [10]. Once again, this definition closely parallels the definition of PET drug in section 121(a) of the FDAMA [5].

It seems that the inclusion of these “ancillary” components and equipment (e.g., nuclide generator, accelerator, electronic synthesizer, etc.) is overly broad and not appropriate within the practical and technical definition of a drug. However, since these items were included in definition of “compounded PET drug” as stated in the section 121 (a) of the FDAMA [5], FDA felt that the definition of “PET drug” in the final rule should include these items as well [7]. Thus, these “ancillary” items are also subject to the same CGMP requirements for PET drugs as stipulated in part 212 [10].

2.3 PET Drug Versus PET Drug Product

Part 212 defines “PET drug product” as a finished dosage form of a PET drug, and it is included in the FDA definition for “PET drug” [10]. As such, the CGMP requirements for PET drugs do apply to PET drugs, as well as PET drug products.

2.4 Compounded and Noncompounded PET Drugs

The FDAMA does not have separate regulations for compounded PET drugs and noncompounded PET drugs [5]. In fact, section 121(b) of the FDAMA states that 4 years after the date of enactment of the FDAMA or 2 years after the FDA establishes approval procedures and CGMP requirements for PET drugs, whichever is longer, the CGMP requirements established by the FDA for PET drugs will apply to compounded or noncompounded PET drugs [5]. In other words, once the aforementioned condition is met, all PET drugs (including compounded PET drugs) would be subject to PET drug CGMP requirements.

During either of these two time frames, section 121(b) of the FDAMA indicated that a compounded PET drug was not adulterated if it was compounded, processed, packed, or held in conformity with the PET compounding standards and official monographs of the United States Pharmacopeia (USP) [5].

2.5 PET Drug Production Versus Practice of Pharmacy

Although part 212 defines “production” of a PET drug as the manufacturing, compounding, processing, packaging, labeling, reprocessing, repacking, relabeling, and testing of a PET drug [10], FDA has traditionally deferred the dispensing of a patient-specific dose and the use of the drug product to state and local authorities. As such, these postproduction activities are regarded by the FDA as the practice of medicine and pharmacy. Consequently, a guidance – a version marked on the front cover with Small Entity Compliance Guide – issued by the FDA in 2011(Guidance) stated that production includes all operations only to the point of final release of a finished dosage form [12]. In general, a routine FDA inspection to ensure compliance with CGMP would focus on activities up to and including the point of final release of a PET drug product [12].

3 CGMP for PET Drugs

CGMP is a minimum standard that ensures that a drug meets the requirements of safety, as well as it has the identity, strength (or potency), quality, and purity characteristics as it is represented to possess. CGMP is demonstrated through written documentation of procedures and practices.

After visiting several PET drug production sites, FDA concluded that a PET drug producer’s status as either a not-for-profit or for-profit entity does not and should not have any significant bearing on the safety, identity, strength, quality, and purity of PET drugs that it produces and distributes for administration to patients. The same conclusion goes to the methods, facilities, and controls that a PET drug production facility needs to ensure product’s safety, identity, strength, quality, and purity. Instead, FDA believes production, and CGMP differences among PET drug producers are primarily a function of the size, scope, and complexity of their production operations.

Consequently, FDA approach to the CGMP regulation for PET drugs has been shaped largely by the statutory and the imperatives of product’s safety, identity, strength, quality, and purity, rather than commercialization or reimbursement concerns. In fact, FDA found that implementing certain production standards and controls could further ensure the production of suitable PET drugs, regardless of differences among the various PET drug production facilities. FDA believes that the welfare of a patient undergoing a PET scan should not depend on where a particular PET drug was produced.

3.1 Unique Aspects of PET Drug Production

PET drug manufacturing procedures differ in a number of important ways from those associated with the manufacture of conventional drug products, mainly due to the short half-lives of PET radioisotopes. These unique aspects are as follows:

• Because of the short physical half-lives of PET radioisotopes, prolonged manufacturing time significantly would erode the useful clinical life of PET drugs. Likewise, the produced PET drug product is administered to the patient usually within a matter of hours. Due to a relatively small number of patients per day, PET drug production facilities generally manufacture the products in response to daily demand. Thus, a maximum of only a few lots are manufactured per day, with one lot equaling one multiple-dose vial.

• Because one lot equals one multiple-dose vial containing a homogeneous solution of a PET drug product, results from end product testing of samples drawn from the single vial have the maximum possible probability of being representative of all the doses administered to patients from that vial. This special characteristic of the produced PET drug prevents sampling or testing error.

• An entire lot may be administered to one or several patients, depending upon the radioactivity remaining in the container at the time of administration. Consequently, the administration of the entire quantity of a lot to a single patient should be anticipated for every lot manufactured. This is an important consideration when establishing the testing limits for certain attributes such as endotoxins and impurities.

• PET drugs usually do not enter a general drug distribution chain. Rather, the entire lot (one vial) is usually distributed directly from the PET drug production facility either to a single medical department or physician for administration to patients or to a nuclear pharmacy for dispensing. Distribution may occur to other PET centers when the geographic proximity will allow for distribution and use within the PET drug product’s half-life parameters.

• The quantities of radioactive active ingredients contained in each lot of a PET drug generally vary from nanogram to milligram amounts, depending upon various product parameters.

3.2 Part 212 Versus Parts 210 and 211

In consideration of these unique natures of PET drugs and PET drug production, FDA recognized that application of certain provisions of the CGMP regulations in parts 210 and 211 [13, 14] to the manufacture of PET drugs might result in unsafe handling or be otherwise inappropriate [7]. Thus, part 212 differs in many significant ways from the CGMP requirements for non-PET drugs found in the regulations of parts 210 and 211[12–14].

Included among these differences are the following (to be described in more details later):

• Fewer required personnel with fewer organizational restrictions consistent with the scope and complexity of operations

• Allowance for multiple operations (or storages) in the same area as long as organization and other controls are adequate

• Streamlined requirements for aseptic processing consistent with the nature of the production process

• Streamlined quality assurance (QA) requirements for components

• Self-verification of significant steps in PET drug production consistent with the scope and complexity of operations

• Same-person oversight of production, review of batch production and control records (please refer to the section titled “Production and Process Controls” for the distinction between master production and control record and batch production and control record), and authorization of product release consistent with the scope and complexity of operations

• Greater flexibility in approaches to determining whether PET drug products conform to their specifications

• Specialized QA requirements for PET drugs produced in multiple sub-batches (sub-batch means a quantity of PET drug having uniform character and quality, within specified limits, that is produced during one succession of multiple irradiations, using a given synthesis and/or purification operation) [10]

• Simplified labeling requirements consistent with the scope and complexity of operations

3.3 Application of Part 212

The regulations in part 212 apply only to the production, quality control (QC), holding, and distribution of PET drugs [10]; human drugs that do not meet the definition of a PET drug must be manufactured in accordance with the CGMP requirements in parts 210 and 211 [13, 14]. Part 212 regulations are designated to be sufficiently flexible to accommodate the unique aspects of PET drugs [10].

However, section 212.5(b) also gives producers of PET drugs under the review of an investigational new drug application (IND) or under a Radioactive Drug Research Committee (RDRC) the option of following the CGMP regulations in part 212 or producing PET drugs in accordance with General Chapter <823> (USP <823>) of the 32nd edition of the USP Radiopharmaceuticals for Positron Emission Tomography – Compounding [10, 15].

Originally developed in 1990, USP <823> sets forth requirements for PET drug production, including control of components, materials, and supplies, verification of procedures, stability testing and expiration dating, QC, and sterilization and sterility assurance [15]. The 1997 FDAMA stipulates that a compounded PET drug is adulterated unless it is produced in conformity with the USP compounding standards (including USP <823>) and official monographs for PET drugs [5].

The main reason that the FDA allows USP <823> to constitute CGMP standards for IND and RDRC PET drugs is to “allow more flexibility during the development of these drugs” [12]. Although the provisions in USP <823> [15] are generally less specific and explicit than the requirements in part 212 [10], FDA believes that provisions in USP <823> are “adequate to ensure that investigational and research PET drugs are produced safely under appropriate conditions” and are “appropriate CGMP requirements for the investigational and research stage of development” [12]. FDA also recognizes that the majority of these investigational and research PET drugs do not have commercial potential [12], and, therefore, it is not necessary to subject these drugs under the stipulations of part 212, which is more geared to approved or to-be-approved PET drugs [10].

Better compliance outcome is another reason that FDA permits producers of IND and RDRC PET drugs to choose USP <823> as an alternative standard for meeting CGMP requirements [12]. FDA stated that because “most PET drug producers are very familiar with the requirements in Chapter <823>, allowing producers to meet the CGMP requirements for investigational and research PET drugs by following Chapter <823> should greatly facilitate producers’ compliance with the CGMP requirements” [12].

Nevertheless, once a PET drug producer intends to seek marketing approval for a PET drug or a new indication, the production of the PET drug to be used in phase 3 study should be conducted in accordance with the CGMP requirements for PET drugs as stipulated in part 212 which is mainly designated as CGMP requirements for approved PET drugs [8].

3.4 QA and QC

On the initiative taken by the FDA, the term “quality control” was replaced with “quality assurance” [8]. Nevertheless, the term “quality control” still appears in the definition of the term “batch production and control record” [10]. The Guidance lists the term “quality control” in the document seven times [12]. Please refer to the section titled “Revised USP <823>” for the definitions of “quality control” and “quality assurance.”

3.5 Personnel and Resources

Section 212.10 requires a PET drug production facility to have a sufficient number of personnel with the necessary education, background, training, and experience to enable them to perform their assigned functions correctly [10]. Each site also must provide adequate resources, including equipment and facilities, to enable their personnel to perform their functions. This section only addresses personnel issues whereas the resources aspects (facilities and equipment) are discussed in section titled “Facilities and Equipment.”

A small PET drug production operation is not properly characterized in part 212 [10]; however, it is stated in the Guidance as one that produces only one or two batches each day (or week) of a single PET drug [12]. As such, it may be adequate to employ only a few employees (typically at least two persons) to accomplish all production and QA functions. The Guidance further elaborates that one individual can be designated to perform the production as well as QA functions, provided this person is highly qualified in the performance of all such functions [12]. Self-checks in such a PET drug production facility are also permissible as indicated in the Guidance [12]. This is not allowed under current part 211 in which second-person checks are required at various stages of productions as well as test verification [15].

3.6 QA

Section 212.20 requires PET drug production facilities to establish and follow written QA procedures for production operations, materials, specifications and processes, and production records to ensure drug.

The Guidance recommends each PET drug production facility to establish a QA function which consists of execution and oversight activities related to QA requirements [12]. It further suggests the following activities to be handled by the execution and oversight of QA function, respectively [12]:

Execution activities of QA function include the following:

• Examine and evaluate each lot of incoming material before use to ensure that the material meets its established specifications.

• Review the production batch records and laboratory control records for accuracy, completeness, and conformance to established specifications before authorizing the final release or rejection of a batch or lot of a PET drug.

• Ensure that deviations from normal procedures are documented and justified.

Oversight activities of QA functions include the following:

• Approve procedures, specifications, process, and methods.

• Ensure that personnel are properly trained and qualified, as appropriate.

• Ensure that PET drugs have adequately defined identity, strength, quality, and purity.

• Ensure that all errors are reviewed. When it is determined that an investigation is appropriate, document the investigation and take corrective action(s) to prevent the recurrence of the errors.

• Conduct periodic audits to monitor compliance with established procedures and practices.

For PET drug production facilities currently producing one or two PET drugs, the Guidance indicates that employees located at the facility can perform both the daily execution and oversight functions [12]. On the other hand, a PET drug production organization managing multiple production facilities may designate the oversight duties to an internal QA department or an outside consultant to achieve a more objective and efficient management [12].

3.7 Facilities and Equipment

3.7.1 Facilities

Section 212.30(a) requires a PET drug production facility to have adequate facilities to ensure the orderly handling of materials and equipment, the prevention of mix-ups, and the prevention of contamination of equipment or product by substances, personnel, or environmental conditions that could reasonably be expected to have an adverse effect on product quality [10].

For the potential sources of contamination, the Guidance points out that they include particulate matter, chemical, and microbiological materials [12]. The general principles for preventing mix-ups as suggested by the Guidance are equipment to be appropriately located, work areas to be organized and proximally located, components to be properly labeled and kept separate according to the classifications (i.e., quarantined, approved, or rejected), and access these aforementioned items to be restricted to authorized personnel [12].

Section titled “Facility Design and Environmental Control” in USP General Chapter <797> Pharmaceutical Compounding – Sterile Preparations is a good reference for a proper setup of an aseptic processing area [16]. Some precautions suggested by the Guidance should be followed to help maintain the appropriate air quality of the aseptic workstation [12]:

• Sanitize the aseptic workstation before each operation.

• Keep items within a laminar airflow aseptic workstation to a minimum and do not interrupt the airflow.

• Have operators wear clean lab coats and sanitized gloves when conducting an aseptic manipulation within the aseptic workstation.

• Frequently sanitize gloved hands or frequently change gloves when working in the aseptic workstation. Examine gloves for damage (tears or holes) and replace them if they are damaged.

• Sanitize the surface of non-sterile items (e.g., test tube rack, the overwrap for sterile syringes, filters) and wipe with an appropriate disinfectant (e.g., sterile 70 % isopropyl alcohol) before placing them in the aseptic workstation.

3.7.2 Equipment

Section 212.30(b) requires the implementation of procedures to ensure that all equipment that would reasonably be expected to adversely affect the identity, strength, quality, or purity of a PET drug, or give erroneous or invalid test results when improperly used or maintained, is clean, suitable for its intended purposes, properly installed, maintained, and capable of repeatedly producing valid results [12]. Activities in accordance with these procedures must be documented [12]. Section 212.30(c) stipulates the equipment to be constructed and maintained so that surfaces that contact components, in-process materials, or drug products are not reactive, additive, or absorptive so as to alter the quality of the PET drug [10].

For any newly installed equipment, the Guidance recommends a three-step process – installation qualification (IQ), operational qualification (OQ), and performance qualification (PQ) – to qualify the new equipment before its first use in order to ensure it was installed correctly, was operated adequately, and is capable of producing the anticipated results [12].

Normally, the equipment vendor verifies that the equipment is installed correctly (IQ) and operates according to specifications, OQ. Before the equipment is used for production, personnel in the PET drug production facility should verify that the equipment, when operated under actual production parameters or selected method, produces consistent results within established specifications – PQ [12].

Representative equipment is discussed in the Guidance to illustrate how it should be controlled [12]. Some examples are listed below:

• The provisions contained in the USP General Chapter <1015> Automated Radiochemical Synthesis Apparatus can help ensure proper functioning of a synthesis apparatus [17].

• Microbiological monitoring (e.g., using settle plate) in the aseptic workstation should be conducted during sterility testing and critical aseptic manipulation.

• In order to make sure that a gas chromatograph (GC) or a high-performance liquid chromatograph (HPLC) system is functioning correctly, a suitability testing should be conducted. At least one injection of the standard preparation (reference standard or internal standard) should be done before the injection of test samples into a GC or HPLC system.

3.8 Control of Components, Containers, and Closures

Section 212.40(a) and (b) requires PET drug production facilities to establish, maintain, and follow written procedures for the control (i.e., receipt, log-in, identification, storage, handling, testing, and acceptance and/or rejection) of components, containers, and closures. Appropriate written specifications for the components, containers, and closures must be established [10]. Section 212.40(c) establishes the minimum standards for controlling components, containers, and closures from receipt to consumption [10]. Any incoming lot must be appropriately designated as quarantined, accepted, or rejected. One must use a reliable supplier for each component, container, and closure [10]. Section 212.40(d) requires that components, containers, and closures be handled and stored in a manner that prevents contamination, mix-ups, and deterioration [10]. Section 212.40(e) requires that PET drug production facilities keep a record of each shipment of each lot of components, containers, and closures that they receive [10].

The Guidance defines “reliable vendor” as “qualified vendors,” and a vendor is qualified when there is evidence to support its ability to supply a material that consistently meets all quality specifications [12]. It is preferable as suggested by the Guidance to have more than one qualified vendor for a component [12].

For acceptance testing, the specific identity test is only required if the finished-product testing does not ensure that the correct components have been used or if an inactive ingredient (e.g., 0.9 % sodium chloride solution) was prepared on site from a component that is not subject to finished-product testing [10, 12]. Otherwise, examination of a certificate of analysis (COA) provided by the supplier should be sufficient [10].

3.9 Production and Process Controls

Section 212.50 requires adequate production and process controls to ensure consistent production of a PET drug that meets the applicable standards for identity, strength, quality, and purity [10].

3.9.1 Master Production and Control Record and Batch Production and Control Record

Section 212.50(b) requires PET drug production facilities to have master production and control records that document all steps in the PET drug production process [10]. Master production and control record (or master record) is a principal document that describes how a drug product is made. It is used to derive an individual batch record (i.e., batch production and control record) and specifies how each batch is to be produced. Since the master record is a compilation of detailed step-by-step instructions, it has been suggested, during the comment period for the final rule, to change the term “control record” to “control procedure” [7]. However, FDA did not accept it because “control record” is a standard term used in the production of drugs [7].

Section 212.50(c) requires that a batch production and control record (i.e., batch record) be created for each new batch of a PET drug [10]. The batch record provides complete traceability and accountability for production and control of each batch of drug product.

3.9.2 Radiochemical Yield Limit

Section 212.50(b)(1) through (b)(6) listed certain items of information that would be required in a master record [10]. One of these required information, as described in section 212.50(b)(6), is a statement of acceptance criteria on radiochemical yield [10]. In the 2005 proposed rule, section 212.50(b)(6) described this statement as acceptance criteria on radiochemical yield (i.e., the minimum percentage of yield beyond which investigation and corrective action are required) (emphasis added) [7].

During the comment period of the 2005 proposed rule, one comment suggested the deletion of this requirement [7]. The comment indicated that radiochemical yields could have significant variations even in a well-controlled PET manufacturing operation and that many factors could affect the yield [7]. The comment maintained that radiochemical yield is not a significant predictor of product quality [7]. According to the comment, discarding useful quality product and having to produce another lot based on arbitrary radiochemical yield increase radiation exposure [7].

FDA’s response to this comment was that although a low radiochemical yield would not necessarily require the rejection of a batch, low radiochemical yield can be a useful predictor of control of the production process for a PET drug [7]. FDA further articulated that a low radiochemical yield might result from a leak in the production system that introduces an extraneous substance, resulting in a contaminated product that might not be easily purified [7].

As such, FDA concluded that repeated occurrences of low radiochemical yield or a downward trend in radiochemical yield should prompt an investigation and, if necessary, corrective action [7]. Nevertheless, FDA did revise section 212.50(b)(6) to require a statement of action limits, rather than acceptance criteria, on radiochemical yield, because not meeting the radiochemical yield limits would require investigation and corrective action but not necessarily rejection of the batch [10]. This is one good example to illustrate the openness and willingness of FDA to work with the PET community during the establishment of PET drug CGMP regulations.

Another example – one comment suggested that it is not necessary to record date and time of each production step. The comment maintained that recording the date once is sufficient since the majority of PET drug productions only take a few hours at most. The comment also recommended that recording the time be limited to critical steps due to the importance of these processes [7]. Although FDA did not accept their suggestion for the removal of date-recording requirement, FDA adopted their recommendation of recording time for each critical production step (e.g., start of irradiation, beginning and end of synthesis) [7, 10].

3.9.3 Process Verification

Section 212.50(f)(2) requires that when the results of the production of an entire batch of a PET drug are not fully verified through finished-product testing or when only the initial sub-batch in a series is tested, the PET drug producer must demonstrate that the process for producing the PET drug is reproducible and is capable of producing a drug product that meets the predetermined acceptance criteria [10]. If each entire batch of a PET drug undergoes full finished-product testing to ensure that the product meets all specifications, section 212.50(f)(1) indicates that process verification as described in section 212.50(f)(2) is not required [10].

For the process verification under section 212.50(f)(2), the Guidance recommends that if a PET drug production facility has an established history of PET drug production, the process verification can be accomplished using historical batch records, provided that there is adequate accumulated data to support a conclusion that the current process yields batches meeting predetermined acceptance criteria [12].

The Guidance also recommends that new processes or significant changes to existing processes be shown to reliably produce PET drugs meeting the predetermined acceptance criteria before any batches are distributed [12]. This verification should be conducted according to a written protocol and generally include at least three consecutive acceptable production runs [12].

Because of the unique characteristics of PET drug production, a PET drug producer may decide to evaluate the reliability of a new process or a significant change to an existing process to produce a PET drug by meeting the predetermined acceptance criteria concurrently with the distribution of the batch, provided that this approach is consistent with the product’s approved application [12]. Such a decision should be justified in writing, subjected to QA procedures, and performed according to a written protocol [12]. Each batch should be processed in strict adherence to the written procedures, fully tested (except sterility testing), and found to comply with all procedural and quality test requirements prior to final release [12].

For computer control, the Guidance indicates that the operation of the computer program be verified before first use or any subsequent changes or upgrades made to the computer program [12]. PET drug production facilities can rely on a certification by the software or system vendor that the specified software was verified under its operating condition [12].

3.10 Laboratory Controls

Section 212.60 requires the establishment and implementation of procedures for testing components, in-process materials, and finished PET drug products [10]. All necessary tests of materials and products must be documented [10]. Each laboratory must have sampling and testing procedures designed to ensure that components, drug product containers and closures, in-process materials, and PET drug products conform to appropriate standards [10]. Analytical methods and test equipment must be suitable for their intended uses [10]. Reagents, solutions, and supplies used in testing procedures must be adequately controlled [10]. The preventive maintenance, calibration, and procedures to make sure that the equipment is functioning properly must be documented [10]. A complete record of all tests related to the production of a PET drug must be kept to ensure compliance with established specifications and standards, including examinations and assays [10].

For reference standards used in analyses, the Guidance states that if a primary reference standard is obtained from an officially recognized source (e.g., USP), acceptance testing of the material is not required [12]. However, if the reference standard is not available from the official source, the identity and purity of the substituted material must be fully confirmed [12].

3.10.1 Stability Testing

Section 212.61 requires the establishment of a written stability testing program for each PET drug product [10]. This program must be used to establish suitable storage conditions as well as expiration dates and times [10].

PET drugs have extremely short shelf lives due to their short half-lives; it might seem to be pointless to evaluate and establish the stability of PET drugs. However, the Guidance indicates that stability concerns for PET drugs are well grounded because of radiation-related radiolysis [12]. As such, appropriate stability parameters such as radiochemical identity and purity (including levels of radiochemical impurities), appearance, pH, stabilizer or preservative effectiveness, and chemical purity should be evaluated to establish and document the stability of PET drug products under proposed storage conditions [12].

Stability testing of the PET drug product should be performed at the highest radioactive concentration, and the whole batch volume in the intended container/closure should be stored [12]. At least three production runs of the final product should be studied for a time period equal to the labeled shelf life of the PET drug product [12].

3.11 Finished Drug Product Controls and Acceptance Criteria

Section 212.70 requires that specifications be established for each PET drug product batch (or sub-batches) including identity, strength, quality, purity, sterility, and apyrogenicity [10]. Except as conditional final release (section 212.70(f)), section 212.70(c) requires an appropriate laboratory determination be conducted to ensure that each batch (or sub-batches) of a PET drug product conforms to specifications, except for sterility, before final release [10].

However, due to the very short half-life of certain PET drug (e.g., ¹³N-ammonia), the release of subsequent sub-batches can be qualified if the initial sub-batch meets all acceptance criteria as per the Guidance [12]. This exception is only permissible provided that a sufficient number of sub-batches (beginning, middle, and end) have been demonstrated to produce a product meeting the predetermined acceptance criteria [12]. In certain cases, it may be appropriate to include testing each sub-batch for certain attributes prior to release (e.g., for pH determination in ¹³N-ammonia production method using Devarda’s alloy catalyst) [12].

3.11.1 Finished-Product Testing

If finished-product testing alone is used, each batch of PET drug should be tested for conformance to all specifications [10]. Otherwise, the Guidance suggests that the following approaches might be employed to achieve the same objective [12]:

• In-process testing – in-process testing might involve the use of an online test to measure an attribute that is equivalent to finished-product testing, provided that the relevant attribute does not change during the production of the finished product.

• Continuous process monitoring of an attribute with statistical process controls – this process involves comprehensive testing of an attribute using online monitoring and corresponding adjustments to prevent an upward or downward drift in batch-to-batch measurements of the attribute.

• Some combination of these approaches:

Although section 212.70(c) addresses conformance to specifications [10], there might be certain QC attributes for a PET drug product that are not as significant as those included in the specifications [12]. Examples of these noncritical QC attributes might include radionuclidic purity (when potentially contaminating radionuclides, do not impact the safety or effectiveness of the drug product), certain low-level nontoxic impurities, and class 3 residual solvents [12, 18].

FDA indicated that a PET drug producer may conduct a periodic quality indicator test (PQIT) on a particular noncritical QC attribute [12]. A PQIT can be performed at predetermined intervals (similar to the skip-testing concept) rather than on a batch-to-batch basis. The selection of noncritical QC attribute(s), as well as the determination of testing frequency, should be set by each PET drug producer to fit within its internal quality system. The PQIT’s QC attribute(s), analytical method(s), limit(s), and frequency should be included in an approved application before implementation.

Under section 212.70(d) when it is determined that all acceptance criteria have been met, the PET drug production facility should then provide a notice of final release to the receiving facility so that the dose may be given to the patient [10].

3.11.2 Sterility Testing

Section 212.70(e) requires that sterility testing be started within 30 h after the completion of PET drug production; however, the 30 h requirement may be exceeded due to a weekend or holiday [10]. The regulation states that if the sample for sterility testing is held longer than 30 h, the PET drug producer must demonstrate that the longer period does not adversely affect the sample and the test results obtained will be equivalent [10]. Tested samples must be from individual batches and not pooled [10]. If the product fails to meet a criterion for sterility, all facilities that received the product must be immediately notified (emphasis added) with the test results [10].

During the public comment period for the 2005 proposed rule, FDA received several comments objected to the proposed requirement to notify receiving facilities immediately if a PET drug product fails the sterility test [6]. The reasons of these comments against the immediate notification include detection of a growth in an inoculated media does not necessarily equate to sterility failure; only confirmed non-sterility rather than technical error should be notified; a sterility test failure might not be known for 2–4 weeks; the observation of any sterility testing failure would occur several days after administration of the drug product and critical data, such as species identification, would not be available; and immediate, unqualified notification would be alarming and unproductive [6].

In FDA response it stated that “[w]e understand that initial results from conventional sterility tests are not definitive, and we appreciate that it takes some time to investigate a failed test. However, we believe that it is important to convey to the clinician the potential risks to a patient when a PET drug product initially fails to meet a criterion for sterility.” [7].

The Guidance recommends that the samples should be stored appropriately (e.g., under refrigeration) [12]. Verification of equivalent results can be accomplished by inoculation of a USP indicator organism (e.g., E. coli) and should demonstrate that there is little, if any, loss in viability of the inoculated microorganism [12].

3.11.3 Conditional Final Release

When one of the required finished-product tests cannot be completed due to a malfunction involving analytical equipment, the product may still be released for human use under the conditions stated in section 212.70(f) [10]. This is another initiative took by the FDA during the establishment of the CGMP requirements for PET drugs [7].

Breakdown Versus Malfunction

Initially the 2005 proposed rule limited the conditional final release to a circumstance of a complete breakdown of analytical equipment [6]. However, the feedback that FDA received during the public comment period indicated that equipment malfunction might be a more appropriate term to depict the circumstance for conditional release [7]. An equipment malfunction might cause atypical test result(s); however, other indicators may show that release of raw materials, production, and purification process events have occurred as expected. For example, a PET drug producer might observe a baseline drift in an HPLC analysis for a product, but if the peak shape is similar to what is normally seen and the production and purification events have progressed as expected, it might be reasonable to conclude that there is an equipment malfunction, rather than that the product is contaminated.

Clearly equipment breakdown as originally mentioned in the 2005 proposed rule [6] might still be an adequate circumstance for meeting the conditional release criteria. In this instance, the term “equipment malfunction” does encompass this situation.

Reserve Sample

Under section 212.70(f)(1)(iii)(iv), which was another FDA’s initiative, a reserve sample of the conditionally released batch of drug product should be retained [10]. The malfunction of analytical equipment should be promptly corrected, and then the omitted test should be completed using the reserve sample [10].

Public comments that FDA received on this particular subject disagreed with this requirement [7]. They argued that, depending on when analytical equipment is repaired, the PET drug producer might not be able to obtain meaningful data for testing (e.g., radionuclidic identity or purity) because the radioactivity of the radionuclide might be decayed to background level [7]. In the response [7], FDA was in agreement that some critical tests would not able to be performed at a later time (i.e., after correction of an analytical equipment malfunction) because of the short half-life of a product.

However, FDA did not think that radionuclidic identity, of which a dose calibrator is required for testing, would be impacted. If the dose calibrator is not functioning properly, FDA indicated that the dose of the product would not be accurately measured [7]. Even if a dose calibrator is malfunctioning and the activity of a product could not be assayed, a sample of known dilution could still be counted using other equipment, and the activity concentration could be determined by correcting for counting efficiency and dilution [7]. As for radionuclidic purity, FDA believed that it is possible to conduct the test on a decayed sample of the product [7].

Overall, FDA recommended that PET drug producers develop alternate tests for specifications for which they conclude it is not possible to conduct a particular test after an analytical equipment malfunction has been corrected [7].

Notification of an Incomplete Testing or an Out-of-Specification Result

Initially the immediate notification was required when the incomplete product testing is noted [6]. Several comments indicated that the personnel at the receiving facility might not be knowledgeable of the conditional final release allowance and lack the expertise to interpret the meaning of such a release in the context of patient safety and product efficacy [7]. The comments stated that the notification in these circumstances would cause uncertainty and undue apprehension, which would not serve the best interest of patients [7]. FDA agreed with these comments and subsequently deleted this requirement from the section 212.70(f)(1) [7, 10].

Nevertheless, section 212.70(f)(1)(iv) requires the PET drug producer to immediately notify the receiving facility if an out-of-specification (OOS) result is obtained when testing the reserve sample (emphases added) [10]. Notifying receiving facilities of OOS results so that personnel can take appropriate action, usually to prevent administration of the drug, is consistent with the intent of CGMP to ensure that patients receive appropriate PET drugs. This differs from the conditional release situation involving notification of incomplete product testing under proposed section 212.70(f)(1)(iii), in which it is still possible that the batch may actually conform to specifications and therefore be appropriate for administration to patients [10].

Radiochemical Identity/Purity and Specific Activity

Under section 212.70(f)(2), conditional final release of a PET drug product cannot be permitted if a PET drug producer could not perform a radiochemical identity/purity test on a PET drug product or could not determine the product’s specific activity [10]. The determination of the product’s specific activity is vital in particular for a PET drug product with mass-dependent target localization and/or potential to elicit a physiological effect, where the specific activity limit is quantitatively expressed.

Releasing of Another Batch

Section 212.70(f)(3) stipulates that the PET drug producers may not release another batch of PET drug product until the aforementioned conditions (i.e., correction of the malfunction problem and completion of the omitted test) are met [10].

FDA felt that these changes would not impose a significant additional burden on PET drug producers because many medical facilities that produce and administer PET drugs might be able to obtain PET drugs for their patients from other PET drug producers while they were correcting an equipment malfunction [7]. Clearly this provision does present a patient-care issue when PET drugs are not readily available from other source(s) due to the very short half-lives of these drug products (e.g., ¹¹C- or ¹³N-labeled PET drugs).

Rejection and Reprocessing

Under section 212.71(a), a batch of a PET drug product that fails to meet established specifications must be rejected, and procedures must be established to identify and segregate the product [10]. Section 212.71(b) requires that documentation of the investigation of a nonconforming product includes the results of the investigation and final disposition of any rejected product [10]. Section 212.71(c) requires corrective action be taken for any identified problems to prevent recurrence of a nonconforming product or other quality problems [10].

Under section 212.71(d), a drug product can be reprocessed if procedures stated in the product’s approved application are followed and the finished product conforms to specifications (except for sterility) before final release [10]. When the option for reprocessing is exercised, FDA recommended that the event be documented and conditions be described in a brief deviation report [12]. Examples of reprocessing could include a second passage through a purification column to remove an impurity or a second passage through a filter if the original filter failed the integrity test [12].

3.12 Packaging and Labeling

Because of radiation exposure concern, the Guidance states that it is a common practice to prepare much of the labeling in advance [12]. For example, an empty product vial can be pre-labeled with partial information (e.g., product name, batch number, date) prior to filtration of the radioactive product, and upon completion of a QC test, the outer shielded container can be labeled with the required information (e.g., radioactivity). Alternatively, a string label can be used to label the immediate container provided that there is a way to associate the label with the vial if the label were to come off. Different approaches can be used as long as the approach ensures that the required information is available on the label.

3.13 Distribution

Section 212.90 requires the development of procedures to ensure that shipment will not adversely affect the product [10]. PET drug production facilities must maintain distribution records for PET drug products. The Guidance further recommends that a system be put in place by which the chain of distribution of each batch of PET drug product can be readily determined to permit its recall if necessary [12]. A recall should consist of notifying the receiving facility, the pharmacist, and the patient’s physician, if known [12].

3.14 Complaint Handling

Section 212.100 states that a PET drug product that is returned because of a complaint or for any other reason may not be reprocessed and must be destroyed in accordance with applicable federal and state law [10].

3.15 Records

Section 212.110 requires all records be reasonably accessible to responsible officials of the facility and FDA investigators and all records be kept for at least 1 year from the date of the final or conditional final release of a PET drug product [10]. The Guidance indicates that verification reports should be kept as long as the systems are in use [12].

4 USP <823>

USP <823> provides QA standards for PET drugs that are produced for compounding, investigational, or research purposes in the United States. USP <823> was originally published in 1998 in the US Pharmacopeia 23 – National Formulary 18 (USP 23–NF 18), Eighth Supplement [19]. Monographs for individual drugs and dosage forms as well as general chapter guidelines are also published in USP–NF in different formats (i.e., print, USB flash drive, or online subscription).

Under a long-standing provision of the FD&C [20], drugs recognized in USP–NF must comply with compendial identity standards, as well as compendial standards for strength, quality, and purity, as set forth in applicable USP monographs and General Chapters. This is described in FD&C 501(b) [20]. However, the USP has no role in enforcement, which is the responsibility of FDA and other government authorities.

Although the provisions in USP 32nd <823> [15] are generally less specific and explicit than the requirements in part 212 [10], FDA believes that they are adequate to ensure that RDRC and IND PET drugs are produced safely under appropriate conditions, consistent with section 501(a)(2)(B) of the FD&C [20], and are appropriate CGMP requirements for the RDRC or IND stage of development.

4.1 Flexibility in USP 32nd <823>

Only the major flexible items in USP 32nd <823> [15] are listed here:

• The post-filtration integrity test is not required prior to the release of the batch of ¹⁵O water for human use.

• The unacceptable QC test result after investigation is not required to be notified to the receiving facilities [15].

• Less demanding on microbiologic testing of aseptic workstations – the Guidance stipulates that this testing “be performed during sterility testing and critical aseptic manipulations” [12], whereas USP 32nd <823> requires only that testing be performed periodically (“e.g., weekly”) [15].

• More liberal in verification of the identities of components, containers, and closures, allowing the identities of these items to be verified by “defined procedures, tests, and/or documented certificates of analysis, as appropriate” [15]. Part 212 stipulates that identity testing must be conducted on an active or inactive ingredient if the finished-product testing cannot “ensure that the correct components have been used” [10].

• Unlike part 212 and the Guidance [10, 12], a sterility test is not required to be performed on every batch of PET drug: “After a record of successful sterility tests is established for a particular PET drug, only the first lot prepared each day shall be subject to a sterility test using cultivation methods. However, when a different PET drug is made at the facility or a new lot of sterile components (for example, filter or final product container) is substituted, then the first daily lot of that PET drug is tested for sterility” [15].

4.2 More Flexibility in Part 212 and the Guidance

In general, CGMP requirements as stated in part 212 and its associated Guidance [10, 12] are for approved PET drugs, and the “less specific and explicit” USP 32nd <823> [15] constitutes the minimum CGMP requirements for investigational and research PET drugs [10]. However, it is interesting to note that several provisions in part 212 and the Guidance [10, 12] are more liberal and accommodating than the requirements in USP 32nd <823> [15]; only significant examples are described here:

• Production procedure is not required to be reviewed and verified at a minimum of once a year.

• Copies of outdated production procedures or computer software programs do not need to be retained.

• The evaluations for chemical purity do not have to include analyses for the presence of starting materials, known intermediates, by-products, and known degradation products.

• Terminal sterilization is an acceptable method for producing sterile PET drugs [12].

• Only one injection, rather than five or six replicate injections as per USP General Chapter <621> Chromatography [21], of the standard preparation is required for checking system suitability before injection of test samples into either a GC system or a HPLC system [12]. The standard preparation for the system suitability evaluation can be either the officially recognized reference standard (e.g., USP) or internal standard where an official standard is not available or if a PET drug production facility establishes its own reference standard [12].

• Process verification is not required if (a) the results of the production of an entire batch of a PET drug are fully verified through finished-product testing and (b) established historical batch records show that “there is adequate accumulated data to support a conclusion that the current process yields batches meeting predetermined acceptance criteria” [12].

• “PET drug production facilities can rely on a certification by the software or system vendor that the specified software was verified under its operating conditions” [12].

• Stability parameters required for testing include “radiochemical identity and purity (including levels of radiochemical impurities), appearance, pH, stabilizer or preservative effectiveness, and chemical purity” [12]. However, USP 32nd <823> stipulates that a PET drug under the stability testing “must meet all acceptance criteria” [15].

• The initiation time for the sterility testing is 30 h (FDA guidance 2011) rather than 24 h as required in USP 32nd <823> [15]. Furthermore, this 30-h requirement can be extended because of an intervening weekend or holiday [10].

• To ensure that each batch of PET drug meets its established acceptance criteria (except for sterility) before final release, the Guidance allows an appropriate laboratory determination to include not only the traditional finished-product testing but also in-process testing, continuous process monitoring of an attribute with statistical process controls, and some combination of these approaches [12].

• PQIT – a PQIT is a noncritical QC test performed at predetermined intervals rather than on a batch-to-batch basis [12]. Examples of these noncritical QC attributes might include “radionuclidic purity (when potentially contaminating radionuclides do not impact the safety or effectiveness of the drug product), as well as certain low-level nontoxic impurities and class 3 residual solvents” [12].

• If the required finished-product tests (with the exception of radiochemical identity/purity test and specific activity test) cannot be completed because of a malfunction of analytical equipment, the related PET drug could still be released for human use under the “conditional final release” provision [10].

• Reprocessing of a batch of a PET drug that does not conform to specifications is allowable [10].

4.3 Revised USP <823>

At the time the monographs and general chapters for PET drugs were published by USP in the 1990s, most PET drugs were produced and used within research or medical institutions. Since then, the environment where PET drugs are produced and used has changed significantly. Today, research or medical institutions continue to produce and use PET drugs for investigational and research purposes. In addition, commercial producers supply most PET drugs used in routine diagnostic imaging procedures. PET imaging agents also have attracted the interest of pharmaceutical companies as potential tools to accelerate and reduce the cost of traditional drug-discovery efforts. Finally, numerous efforts are underway to develop new routine diagnostic imaging agents for use in cardiology, oncology, and neurology. Thus, the use of PET imaging agents today spans discovery, research, clinical development, and routine diagnostic imaging procedures. The diversification of PET during the past 15–20 years has resulted in new requirements for the PET environment, including a greater number of PET drugs, higher production levels, shorter synthesis times, shorter quality control times, more complex syntheses, and increased regulatory oversight.

To understand these changes, USP jointly sponsored two symposia with the Society of Nuclear Medicine (SNM) during SNM’s annual meetings in 2008 and 2009. USP staff and members of the USP Expert Committee on Radiopharmaceutical and Medical Imaging Agents (RMI EC) presented talks and led discussions about historical trends and changes in the PET environment. One goal of these symposia was to describe issues related to USP General Chapters for PET drugs and to gather feedback from the PET community. The USP–SNM joint symposia and regulatory requirements led USP RMI EC to conclude that USP 32nd <823> must be revised.

Deficiencies in the USP 32nd <823> [15] were identified as follows:

• Differences in the organization of USP 32nd <823> compared to FDA’s final rule and Guidance on CGMP for PET drugs [10, 12]

• Further enhancement of USP <823>’s flexible provisions by incorporating items such as PQIT, a 30-h initiation time frame for sterility testing, and conditional final release according to FDA’s CGMP requirements for PET drugs [10, 12]

• The need for consistency with other revision efforts for USP General Chapters

• Inappropriate methodologies for system suitability and quantitative analysis in current chromatographic tests

• Lack of defined frequency for certain QC tests

• Lack of discussion about the timing of the completion of certain QC tests relative to product release

• Lack of requirements for OOS investigations for quality control tests

4.3.1 The Revision Process for USP <823>

To address these deficiencies, USP RMI EC proposed the establishment of an Ad Hoc Advisory Panel (Advisory Panel) composed of academic and industrial members of the PET community. The goal of the USP Advisory Panel was to advise RMI EC about suitable revisions to USP 32nd <823> in accordance with USP’s mission. The formation of the USP Advisory Panel was completed in late 2008, and beginning of July 2009, the USP Advisory Panel met or teleconferenced numerous times. The outcome of this effort is summarized here.

4.3.2 Organization of the Revised USP <823>

To reflect the new role of the revised USP <823> in part 212 [10] and to uphold PET compounding practice, the title of the revised USP <823> has been changed to Positron Emission Tomography – Drugs for Compounding, Investigational, and Research Uses [22].

By means of compounding, pharmacists (or other qualified individuals working under the authority and supervision of a physician), according to state-regulated practice of medicine and pharmacy, fulfill an essential health-care need – providing patients with medications tailored to their needs. In some cases compounding pharmacists provide a drug that is not commercially available. In other cases the patient may be allergic to certain ingredient(s) of the drug, or the dosage form may not be suitable for administration to the patient. In addition to giving patients access to otherwise unavailable or more appropriate PET drugs, compounding may also be used for teaching or QC purposes.

In FDA’s view, part 212 covers both compounded and noncompounded PET drugs and thus believes that these two types of PET drugs should be within the scope of FDA’s authority in accordance of part 212 [7]. However, USP is not addressing the extent of FDA’s regulatory authority, and therefore, by its terms, USP <823> should apply to compounding of PET drugs for human use whether such compounding is or is not subject to federal regulation.

To streamline USP General Chapters related to PET drugs, the USP Advisory Panel decided to consolidate key standards and requirements stated in USP <1015> [17] into the revised USP <823>. Consequently, USP <1015> will be deleted from USP. In addition, a new proposed general chapter numbered <1823>, “Positron Emission Tomography Drugs-Information” [23], has recently been developed to supplement <823>. This new general chapter describes concepts, technologies, and procedures for the preparation of PET drugs and QA. These supplemental descriptions are more suitable for an informational general chapter. This is consistent with USP’s General Chapter Design Project, which is an effort to exclude nonenforceable information from general chapters numbered less than 1000 (USP General Chapter Project Team 2009).

The following are the sections in the revised USP <823> [22]:

• Definitions

• Adequate Personnel and Resources

• QA

• Facilities and Equipment

• Control of Components, Materials, and Supplies

• Process and Operational Controls

• Stability

• Controls and Acceptance Criteria for Finished PET Drugs

• If a PET Drug Does Not Conform to Specifications

• Reprocessing

• Labeling and Packaging

4.3.3 Definitions

The definitions discussed below are different from those defined in the final rule:

Batch Versus Lot

Part 212 uses the terms batch and lot interchangeably [10]. In addition, part 212 appears to use lot synonymously with sub-batch [10]. These definitions and usages may be confusing to the PET community. To differentiate the specific meaning of each term, the USP Advisory Panel and RMI EC propose that the definition of batch applies explicitly to the PET drug and that the definition of lot applies only to components used in the preparation (including QC) of a PET drug. As defined in the revised USP <823>, the term lot means a quantity of materials (e.g., reagents, solvents, gases, purification columns, and other auxiliary materials) that have uniform character and quality within specified limits and are used to make a PET drug product [22].

PET Drug

The definition of PET drug in the part 212 [10] includes “any non-radioactive reagent, reagent kit, ingredient, nuclide generator, accelerator, target material, electronic synthesizer, or other apparatus or computer program to be used in the preparation of a PET drug.” USP feels that nonradioactive reagents, reagent kits, ingredients, and target materials are components used to produce a PET drug and that radionuclide generators, accelerators, electronic synthesizers, and computer programs are ancillary items used in the production of PET drugs. As such, the revised USP <823> defines PET drug as “a finished form of a radioactive drug that exhibits spontaneous disintegration of unstable nuclei by the emission of positrons and is intended for human administration in diagnosis or therapy” [22].

FDAMA defines the “compounded” PET drug to be used for diagnostic images [5]. As such, therapeutic PET drugs are explicitly excluded from part 212 [10]. However, USP believes that the revised USP <823> should apply to any PET drug whether it is for diagnostic or therapeutic use. Therefore, the definition of PET drug in the revised USP <823> includes “therapeutic” use (e.g., tumor therapy) [22].

QA Versus QC

QA and QC are commonly used interchangeably in the PET community even though the terms have fundamental differences. As per the revised USP <823>, QA and QC are defined as follows [22]:

• QA: A planned system for ensuring that a PET drug product possesses defined identity, strength, quality, and purity required for its intended purpose by procedures, tests, and analytical methods.

• QC: A system for testing the quality of components, materials, supplies, and PET drug products by procedures, tests, analytical methods, and acceptance criteria.

Thus, QC is a subset of QA that deals with testing materials and products to determine if they meet acceptance criteria.

Strength

Part 212 defines strength as radioactivity on a volume or weight basis [10]. USP believes that this definition risks confusing strength and specific activity because it is not clear if weight refers to the cold mass of the active pharmaceutical ingredient or the overall weight of the solution.

In addition, it is common practice in PET to define strength on a volume basis. Therefore, strength is defined in the revision of USP 32nd <823> strictly based on volume (e.g., mCi/mL or MBq/mL) which is the same as specific concentration.

Validation Versus Verification

Part 212 has a definition of verification; however, it does not define the term validation [10]. Although the terms validate, validated, and validation are used in the Guidance [12], these terms are also not defined there.

Validation and verification are essential and complementary elements in the CGMP process. The revised USP <823> defines validation as the “establishment of documented evidence that a method, process, or system accomplishes its intended requirements” [22]. In addition, verification is defined as “confirmation that an established method, process, or system meets predetermined acceptance criteria” [22]. It is helpful to think of validation as “building the right thing,” and verification as “building it right.”

4.3.4 Adequate Personnel and Resources

Adequate personnel and resources are addressed in several sections of the 32nd version of USP <823> [15], including Compounding Procedure Verification, PET Radiopharmaceutical Compounding for Human Use, and Quality Control. The revised revision of USP <823> includes a separate section titled Adequate Personnel and Resources, which requires a sufficient number of personnel with appropriate education and training and indicates that the number of personnel depends on the size and complexity of the facility [22]. This section of the revised USP <823> reflects the layout of part 212 addressing personnel and resources as the first topic [10, 22].

4.3.5 QA

This section of the revised USP <823> describes the difference between QA and QC in the production of PET drugs [22]. QA covers all matters that influence the product’s identity, strength, purity, and quality. QC is a subset of QA that deals with testing materials and products to determine if they meet acceptance criteria. The QA function typically consists of oversight activities and the QC function consists of execution activities [24]. According to the Guidance, QA function in a PET drug production facility consists of oversight and execution activities [12].

4.3.6 Facilities and Equipment

The revised USP <823> requires the certification (i.e., integrity testing of the high-efficiency particulate air [HEPA] filter) of the aseptic workstation be performed at least annually [22] when the Guidance requires this every 6 months [12]. As for the microbiological monitoring (e.g., using swab or contact/settle plate) in the aseptic workstation, revised USP <823> requires it be assessed after use and each day of use [22], whereas the Guidance requires it be conducted during sterility testing and critical aseptic manipulation [12].

Cleaning Equipment and Components

The 32nd USP <823> describes the cleaning of equipment but does not specifically address cleaning between batches of PET drugs [15]. As a result, there has been confusion about the acceptability of cleaning between batches. The revised USP <823> addresses this deficiency by describing requirements for cleaning equipment between multiple batches or one of more PET drugs [22]. This approach is consistent with the equipment cleaning requirements described in the Guidance [12].

System Suitability for QC Equipment

The Guidance states that at least one injection of a standard is required for system suitability [12]. The Guidance does not address reproducibility as a part of system suitability but instead references USP <621> which describes system suitability requirements (resolution, replicate injections, and tailing factor) [12, 21].

Although these requirements are important for chromatography systems used in PET drug production, the number of injections required for replicate injections in USP <621> (i.e., five injections for reproducibility if the relative standard deviation is 2.0 % or less and six injections if the relative standard deviation is more than 2.0 %) may not be practical because of the nature (i.e., half-life) and number of different PET drugs and/or the number of batches prepared at a typical academic or commercial PET facility [21]. Consequently, the revision of USP <823> describes two acceptable system suitability approaches that can be used for chromatographic systems [22].

The first approach is the construction of a calibration curve that can be used for an extended period of time. In routine use, the injection of a known standard is used to verify that the calibration curve is appropriate for use in subsequent sample injections. The second approach is the use of a single-point calibration created from two injections of a known standard at the beginning of each testing cycle. In each case, the requirement for replicate injections is met by comparison of multiple injections, either within a single testing cycle or over an extended period of time. Together, the approaches described in the revised USP <823> provide more flexibility and clarity than does the Guidance [12, 22].

Other chromatographic parameters such as signal-to-noise ratio, limit of detection, and limit of quantitation can be determined as part of routine system suitability testing [22]. System suitability tests also may be appropriate for other QC equipment, including dose calibrator, radio-thin layer chromatography, and multichannel analyzer [22].

4.3.7 Control of Components, Materials, and Supplies

If growth media used in the sterility testing of PET drug products are obtained from commercial sources, the revised USP <823> requires a growth promotion testing that uses a suitable single species of organism be performed on initial qualification of the supplier and periodically (e.g., quarterly) thereafter [22]. This is not required by part 212 or the Guidance if the manufacturer of the growth media provides COA that contains the results of the growth promotion testing [10, 12].

4.3.8 Process and Operational Controls

This section of the revised USP <823> covers the “Production and Process Controls” and “Laboratory Control” sections as described in part 212 and the Guidance [10, 12, 22].

Acceptance criteria must be established in the master record for each PET drug, and if a USP monograph exists, USP standards are the minimum acceptable requirements unless there are acceptance criteria specified in an FDA-approved IND or RDRC protocol [22].

Aseptic Operation

The revision of USP <823> allows the preparation of multiple PET drug vial assemblies in a single aseptic operation cycle [22]. It also allows the sterility testing with the use of a septum-sealed sterility test tube be carried out in a shielded area that does not contain an HEPA filter [22]. This provision is to address radiation shielding requirements when inoculating radioactive samples for sterility test [22]. USP felt that the flexibility is necessary since many hot cells used for this purpose do not have laminar airflow and the use of a media tube with a septum cap provides adequate protection from accidental contamination during the inoculation process. If media tubes have a screw cap opening, inoculations must be performed in the aseptic workstation. The Guidance does not differentiate inoculation requirements for screw cap or septum cap media tubes [12].

Process Verification

Documented studies must be performed to ensure that the process described in the master record yields a PET drug that meets the defined acceptance criteria.

The revised USP <823> indicates that the process verification must be completed on three batches that do not have to be consecutive [22]. This allowance was added to account for a batch that may not be completed because of factors that are not relevant to the quality characteristics of the PET drug (e.g., cyclotron malfunction, hardware malfunction, and so on). This allows the completion of process validation without the potential losses that could occur with three consecutive batches stipulated in the Guidance [12].

As per part 212, process verification is required if an entire batch of a PET drug is not fully verified through finished-product testing or when only the initial sub-batch in a series is tested [10]. The Guidance allows the use of historical batch records to substitute the process verification provided that there is adequate accumulated data to support the current process that would yield batches meeting the predetermined acceptance criteria [12]. Due to the unique characteristics of PET drug production, FDA allows the distribution of the batch of PET drug under process verification if the produced PET drug met the predetermined acceptance criteria [12].

Based on these statements, USP should consider adding similar requirements in USP <823> to clarify when the process verification is or is not required to be performed.

4.3.9 Stability

The revised USP <823> [22] is consistent with the Guidance [12] in the evaluation parameters and test conditions to establish and document the stability profile of PET drug products under proposed storage conditions.

4.3.10 Controls and Acceptance Criteria for Finished PET Drug Products

The 32nd USP <823> permits a reduction in the frequency of sterility tests after a record of successful sterility tests is established for a particular PET drug [15]. Because sterility testing and bacterial endotoxins test (BET) are both biological assessments, the revised USP <823> includes a reduction in the frequency of the BET [22]. Similar to the sterility test, the BET must be performed on the first batch of each PET drug prepared each day [22].

The 32nd USP <823> describes an in-process 20-min endotoxin test [15]. USP believes that this test is out of date and is too prescriptive (i.e., a 20-min process and “incorporating positive controls in the range of 5 EU per mL to 175 EU/V, where V is the maximum volume of injection”). Therefore, the revision of USP <823> does not include the 20-min in-process test [22].

If a PET Drug Does Not Conform to Specifications

If a PET drug does not conform to specifications, the first action generally is to investigate the QC process. Such investigations typically are known as OOS investigations which are not addressed in the USP 32nd <823> [15].

The guidance for OOS investigations of traditional pharmaceuticals [25] does not apply well to short-lived PET drugs. Section 212.71 addresses the question, “What actions must I take if a batch of PET drug does not conform to specifications?” [10] This section requires the rejection of PET drugs that do not meet specifications but does not discuss OOS investigations and the possibility of analytical error as a cause of OOS results. The result is that none of the FDA documents adequately address OOS investigations for PET drugs. To resolve these shortcomings, the revised USP <823> includes a description of OOS investigations [22].

The focus of an OOS investigation is to determine if the OOS QC finding is the result of an analytical error or a true product failure. If the investigation determines that the OOS result is caused by analytical error, the original test results are invalidated. Thus, an OOS result does not necessarily mean that the batch fails and must be rejected. However, if the investigation determines that the OOS result indicates a true product failure, the batch must be rejected, and the cause of the failure must be investigated. However, the revised USP <823> does not address the notification requirements to the receiving facilities.

When a sterility test for a PET drug product shows signs of microbial growth, revised USP <823> indicates that the test result is an OOS and should be investigated [22]. Upon completion of the investigation, immediately notify all receiving facilities if the product fails to meet the criterion for sterility, including the microbiological findings from the investigation [22]. According to part 212, if the PET drug product fails the sterility test, immediate notification to all receiving facilities must be initiated and provide these facilities with the test results, as well as any appropriate recommendations [10]. Upon completion of an investigation, the findings must be immediately notified to all receiving facilities [10].

4.3.11 Reprocessing

Depending on the nature of the failure, the rejected PET drug product may be reprocessed according to preestablished written procedures [22].

4.3.12 Labeling

Labeling requirements for PET drugs are not described in the 32nd USP <823> but are included in a separate section in the proposed revision [15]. The section in the revised USP <823> is divided into information required on the immediate container for PET drugs and on the immediate shielding during storage and use [22].

The revised USP <823> does not list the “Distribution” and “Complaint Handling” sections [22]. This is probably due to the fact that the revised USP <823> only addresses PET drugs for compounding, IND, and RDRC which are most likely not distributed to outside of the facility. The “Records” section described in part 212 and the Guidance [10, 12] is also not included in the revised USP <823> as the record requirements (e.g., composition and quality of the PET drug product, process and production records, and OOS results, etc.) are placed in the related sections of the revised USP <823> [22].

5 Inspection on the PET Drug Production Facility

FDA inspects manufacturers or processors of FDA-regulated products to verify that they comply with relevant regulations. In general, a routine FDA inspection to ensure compliance with CGMP would focus on activities up to and including the point of final release of a PET drug product.

5.1 Types of FDA Inspections

There are three types of FDA inspections on drug production facilities:

• Preapproval inspection (PAI) – this is for new NDAs and ANDAs. The purpose of a PAI is to verify existence of the raw data included in the submission (often the focus here will be on laboratory results for lot release and stability as well as validation) and confirm that the facility is capable of manufacture (e.g., correct scale qualified equipment, process validation, etc.). Yes, there will be an overall CGMP evaluation, but the stated purpose of the PAI (in response to the generic scandal of the 1980s) is focused on those two items. The outcome is a recommendation for approval or not. If the site has no approved products, FDA cannot issue a warning letter, but they will block approval of the product until such time as remediation occurs and has been confirmed by a reinspection.

• Routine surveillance CGMP inspection – periodically (every 2–3 years generally as resources and priorities allow), the FDA will perform an unannounced CGMP inspection for manufacturers who make commercial products. FDA generally does not inspect clinical manufacturing sites unless there are known issues, such as unexpected deaths in a clinical trial that may be linked to method of manufacture. This can cover all CGMP systems or can be a more limited (abbreviated) inspection.

• Compliance inspection – this is either a follow-up inspection post of a regulatory action (e.g., warning letter, untitled letter, regulatory meeting, etc.) or an inspection on a for-cause basis (i.e., when FDA becomes aware of a potential safety concern related to the production of an investigational or research drug). Similar to inspection of production facilities for non-PET investigational drugs, FDA generally will only conduct for-cause inspections of facilities that produce investigational or research PET drugs.

A full inspection covers at least four items, whereas an abbreviated inspection consists of at least two items of the inspection systems listed below:

• Quality system with aseptic sterility controls

• Facilities and equipment system

• Materials system

• Production system

• Laboratory control system

• Packaging and labeling system

The first item “quality system and aseptic sterility controls” is mandatory for both full and abbreviated inspections.

5.2 Selection of Inspection Coverage

According to an FDA talk titled “PET Drug Inspection,” which was presented at the 2014 Society of Nuclear Medicine and Molecular Imaging (SNMMI) annual meeting [25], in 2013 100 % of all FDA inspections on PET drug production facilities were full inspection (at least four systems to be inspected) conducted due to any of the following circumstances:

• Firm has never been inspected.

• Follow-up to regulatory action.

• Significant manufacturing changes.

• Microbial contamination or cross-contamination.

• Poor compliance history – initial inspection conducted in 2012.

In 2014 (as of June 9, 2014, when this presentation was given), FDA indicated that approximately 10 % facilities might get abbreviated inspection (at least two systems to be inspected) for any of the following reasons [26]:

• Adequate compliance history.

• Firm has been inspected for similar class of product.

5.3 Major Issues Identified in 2013/2014 PET Drug Inspections

This FDA report at the 2014 SNMMI meeting [25] revealed several major deficiency issues during their inspections carried out in 2013 and up to June 2014:

• Lack of appropriate training and QA oversight

• Inadequate aseptic techniques and environmental monitoring

• Analytical assay methods not validated or verified for intended use

• Inappropriate controls of equipment and materials

• Deficiencies in production and process controls (including the inadequate media fill control process)

• Deficiencies in laboratory controls

• Lack of adequate finished drug product controls and acceptance criteria (including lack of proper “out-of-specification” investigations)

• Lack of appropriate record keeping

The more specific deficiencies for each of the abovementioned six major issues are listed as follows [25]:

5.3.1 Deficiencies in Appropriate Training and QA Oversight

Failure to train employees, identify as deviation and conduct investigation, review batch record for release, review records in a timely manner, reject batches with OOS result, maintain on-site software, follow established procedures, document handling of nonconformance, and establish a change control procedure.

5.3.2 Aseptic Operation Deficiencies

Identified deficiencies of aseptic operation are:

• The use of an overcrowded laminar flow hood

• The use of non-sterile disinfectants

• Dilution of drug product in unclassified area

• Frequency of environmental monitoring not reflecting manufacturing operation

• Storage of bulk product vial in non-sterile bag

• Out of limit personnel monitoring results

• No record of aseptic area been cleaned

• Lack of documentation on the qualification of personnel conducting aseptic operation

5.3.3 Inappropriate Controls of Equipment and Materials

• Inadequate qualification of chemical synthesizer

• Cleaning not adequate for multiproduct equipment

• Failure to fully evaluate all materials received for use in production

5.3.4 Deficiencies in Production and Process Controls

These deficiencies in production and process controls are:

• Master record does not contain complete instruction of production, sampling, and testing (use of sticky notes).

• Written procedures not specific enough (corporate standard operation procedure).

• No determination to conduct investigation when errors occurred.

• Deficiencies in batch record:

  • Failure to record manufacturer’s name and/or expiration dates of material utilized in production

  • Does not include actual weights or volumes of the components used in the process

  • Ensuring all portions are complete before final quality approval

• Verification not performed to ensure consistency of the process upon process change

• Revalidation not done on computer-based programmable logic controller

5.3.5 Media Fills Deficiencies

Media fills did not simulate production: bulk product vial hold was not the worst case, dilution step not simulated, manufacturing process flow not matched, and withdrawal of QC samples not simulated, and there was no positive control and negative control. Also, media does not come into contact with all interior surface, and not all production personnel participate in media fills.

5.3.6 Deficiencies in Laboratory Controls

The reported deficiencies in laboratory controls are listed below:

Analytical Method Validation

Failure to validate methods which deviate from USP test method

Analytical Equipment Qualification

Failure to qualify equipment for intended use and failure to maintain and calibrate equipment – dose calibration not calibrated for linearity to cover the range being test, multichannel analyzer not recalibrated, and thin-layer chromatography scanner. No audit trail for GC, HPLC, and allowed ability to delete raw electronic test data

System Suitability

GC and HPLC – number of injections: one vs. three. Lack of documentation on daily system suitability. Deficient in analytical balance accuracy verification and pH indicator range

Expiry of Reagent, Reference Standards

No justification for expiration dating of ¹⁸F-fludeoxyglucose reference solution, Kryptofix, buffers, and mannose triflate

5.3.7 Lack of Adequate Finished Drug Product Controls and Acceptance Criteria

Several deficiencies in this area were releasing drug that did not pass specification, releasing without calculating or reporting the radiochemical purity in the batch record, failure to conduct all QC tests for release, and releasing multiple lots under conditional release.

Lack of Proper “Out-of-Specification” Investigations

OOS test averaged to generating passing results and not retaining raw test data.

5.3.8 Lack of Appropriate Record Keeping

Failure to retain raw analytical data, no documentation that sterility test was conducted and the lot was released and equipment maintenance log missing sufficient details