ININ THETHEFIELD FIELD Pharmaceutical Science & Technology News

GMPs Small Changes,Big Effects in Biological Manufacturing

European Union-mandated formu- incidence of PRCA per 100,000 patient response to the recombinant human ery- lation change—intended to increase years as follows: Eprex without HSA, 18; thropoietin. These antibodies also attacked A the safety of Johnson & Johnson’s Eprex with HSA, 6; NeoRecormon, 1; the body’s own erythropoietin, effectively Eprex (injected erythropoietin)—appears Epogen, 0.2. halting almost all signals for the bone mar- to have caused a 10- to 20-fold increase in J&J initiated a 100-person crash pro- row to make red blood cells. The result the incidence of a rare but serious com- gram to locate and correct the cause of the was PRCA, a severe anemia requiring plication among recipients of the drug. problem. At October’s Biotech 2004 meet- transfusions. Recent reports describe the origin of the ing in Philadelphia, Thomas S. Temple- In 2002, J&J switched to PTFE-coated adverse reactions and the detective work man, director of biotechnology develop- rubber stoppers, which have halted the that found and fixed the source of the ment at J&J’s Global Biologics Supply leaching problem and greatly reduced im- problem. Chain, LLC, reported the results of that mune response. The company now cites In 1998, the EU directed Johnson & effort. The polysorbate 80, added to re- its experience as a strong argument for re- Johnson (J&J, New Brunswick, NJ, place the human serum albumin, inter- quiring new clinical trials whenever a new www.jnj.com) to stop using human serum acted with the uncoated rubber stoppers process begins to produce even tried-and- albumin (HSA) as a stabilizer in Eprex, long used in single-use Eprex syringes. The true biologicals—as would be the case the recombinant erythropoietin (EPO, a stoppers leached small amounts of plas- with producers of follow-on biologicals. red blood cell growth factor) sold outside ticizers into the drug solution. These –Douglas McCormick the United States. Polysorbate 80 replaced leachates acted as an adjuvant, stimulat- HSA in the formulation. ing a very small percentage of patients to 1 C.L. Bennett et al. “Pure Red-Cell Aplasia and Epo- Over the next six years, dialysis centers, mount a strong immunoglobulin G etin Therapy.” N.Engl.J.Med. 351, 1403–1408 (2004). first in France and then world-wide, began to notice an increased rate of a serious ane- REGULATORY mia—pure red-cell aplasia (PRCA)— among patients receiving Eprex. The ini- tial reports included 13 patients (11 FDA Publishes Final PAT Guidance receiving Eprex and two receiving Neo- Recormon, an erythropoietin beta mar- The Food and Drug Administration laboratory testing of collected samples to keted by Roche). Further investigation— has published the final guidance on evaluate quality. PAT involves systems for reported in the 30 September New England process analytical technology (PAT). The designing, analyzing, and controlling pro- Journal of Medicine1 by a team of re- document, Guidance for Industry: Process cessing through timely measurements— searchers from the Jesse Brown Veterans Analytical Technology—A Framework for in-line, on-line, or at-line—of critical Affairs Medical Center (Chicago, IL), Innovative Pharmaceutical Development, quality attributes. The guidance also es- Northwestern University (Chicago, IL), Manufacturing, and Quality Assurance, was tablishes a regulatory pathway for imple- Inserm (Paris), and several other institu- posted on the agency’s Web site on 29 Sep- menting this new approach with the aim tions in the US and Western Europe— tember (www.fda.gov/cder/guidance/ of alleviating industry concern that such turned up 191 cases of PRCA with onset 6419fnl.htm) and announced in the Fed- innovation will lead to a regulatory between January 1998 and April 2004: 175 eral Register on 4 October. impasse. associated with Eprex, 11 with NeoRecor- The PAT guidance describes a regula- The final PAT guidance was one of sev- mon, and 5 with Epogen, the HSA-con- tory framework designed to encourage im- eral documents issued at the end of Sep- taining erythropoietin marketed by provement in pharmaceutical develop- tember, marking the two-year anniversary (Thousand Oaks, CA). Incidence ment, manufacturing, and quality of FDA’s initiative,“CGMPs for the Twenty- of drug-associated PRCA was similar from assurance through innovation in product First Century: A Risk-Based Approach” (see region to region, so most of the cases ap- and process development, process analy- “Washington Report,”page 28). peared in the largest markets: France, sis, and process control. Conventional Although most of the changes to the Canada, the UK, and Spain. The re- pharmaceutical manufacturing generally guidance were minor rewordings and clar- searchers estimated exposure-adjusted relies on batch processing with off-line PAT continued on page 18

16 Pharmaceutical Technology NOVEMBER 2004 www.pharmtech.com

IN THE FIELD

PAT continued from page 16 is not my job. Look at the guidance. It says ment. In the current state, continuous im- ifications, a few substantive changes ap- it’s for manufacturing.’The final guidance provement is not possible.” peared. The most significant of these ex- clarifies that PAT is intended to be used any- Stephen Closs, a senior pharmaceutical panded the scope of the guidance to in- where it can provide value.” process development engineer at Patheon, clude the Center for Drug Evaluation and Colin Minchom, Canadian vice-presi- sees this connection between PAT and Research’s Office of Biotechnology Prod- dent of pharmaceutical development serv- quality systems. “The links to ICH guid- ucts, which was created after the draft PAT ices at Patheon Inc. (Mississauga, ON), ances Q8 and Q9 offer opportunities for guidance was issued. agrees that the mention of pharmaceuti- industry to develop quality into processes,” Another important change was that the cal development is important, but empha- he says, referring to the International Con- section on process understanding was sizes a different benefit.“By using PAT, you ference on Harmonization’s Q8 guidance, moved forward to emphasize the guid- can create tools to speed up the develop- Pharmaceutical Development—Quality by ance’s focus. “The main emphasis of PAT ment process,” he says, because on-line Design, and Q9, Risk Management. Hus- is understanding and controlling the tools can allow you to see the conse- sain points out that the approach is now process, and not trying to find exotic tech- quences of scientific decisions faster.“We international. “ICH Q8 will bring all the nologies for doing so,”says Ajaz Hussain, see that as a distinct advantage.” principles of PAT within that framework PhD, deputy director of FDA’s Office of Hussain stresses that PAT fits into a of quality systems,” he says. “So this is a Pharmaceutical Science. “Novelty for the broader concept of quality systems, a po- harmonized approach coming forward.” sake of novelty is not what we are looking sition that the agency detailed in a white Industry members also felt that some for. A focus on technology or sensors, paper, “Innovation and Continuous Im- of the apparently minor editorial changes without really understanding what their provement in Pharmaceutical Manufac- were significant.“Nowhere in the new doc- role is, is a waste of time and actually in- turing.” Issued the same day as the PAT ument do they refer to a specific technol- creases cost.” guidance, the white paper is a summary ogy,”says Mathis. “That’s very positive. It The final PAT guidance also encourages of learnings from the CGMP initiative and embraces the concept that new technolo- the use of PAT in product development. proposed next steps for moving to what gies will come out and that a company “This should end the finger pointing,”says the agency calls the “desired state” of phar- shouldn’t have to look to the guidance to Nancy Mathis, PhD, president and CEO of maceutical manufacturing. see whether a particular technology is Mathis Instruments Ltd. (Fredericton, NB). “What people have to understand is that listed or not.” Hussain agrees.“We have al- “When I talk to people in the industry, the the paradigm has shifted,” says Hussain. ways maintained the position that we will manufacturing people say, ‘PAT has to be “The entire approach to process valida- not recommend a particular technology,” built in and come over in a tech transfer,’ tion and specifications now must be he says. “It’s up to companies to see what which ultimately points back to formula- viewed in the context of quality systems. is applicable and appropriate for a given tion. But the people in formulation say,‘PAT And it is a way to continuous improve- process or product.” Mathis, who is a member of ASTM’s Committee E55 on the Pharmaceutical REGULATORY Application of PAT, noted that the defini- tions of on-line, at-line, and in-line now Aseptic Processing Guidance Is Final match those in the standard on terminol- ogy issued by ASTM in May. “That kind Also on 29 September,FDA published its which summarizes clean air classifica- of harmonization is important so that final guidance, Sterile Drug Products Pro- tions and recommended microbial ac- people don’t get confused,” she says. The duced by Aseptic Processing—Current tion levels, has been modified to ac- final guidance also makes direct reference Good Manufacturing Practice. This com- knowledge that alternative action lev- to standards being established by commit- pletes the agency’s effort to replace the els can be justified, depending on the tee E55. 1987 industry guidance of the same title. method of analysis used.Clarifications The final guidance also removes almost FDA published a revised draft in Septem- also have been made regarding all references to the application of PAT to ber 2003,based in part on recommenda- process simulations. In addition, the specific processes such as blending, an- tions from an aseptic processing working guidance recommends “building other change that Mathis likes. “It opens group formed under the Product Quality quality into products” and under- people’s minds that PAT can be applied to Research Institute. scores the agency’s “encouragement other processes such as those taking place Major changes from the draft include of alternative approaches and innova- in a liquid mixing tank, a reactor vessel, the revision of the sterility testing section tions to achieve increased sterility or a fluid-bed dryer, for example,”she says. to clearly emphasize and reference US assurance.” “Before I think people were thinking too Pharmacopeial Sterility Test ͗71͘.Table 1, -Laura Bush much about blending.” –Laura Bush

18 Pharmaceutical Technology OCTOBER 2004 www.pharmtech.com

PACKAGING Accenture Releases Results of RFID Test Program

After two months of testing and a year The program involved the tracking of and [build] a safe and secure environ- of designing, Accenture (Chicago, IL, 10 real products through 15 locations. ment in terms of track and trace of prod- www.accenture.com) has released the re- Manufacturers were required to verify read- uct at the item level—from the manufac- sults of its radio frequency identification ability of the tags before applying them to turer to point of dispensing,” says Jaime (RFID) prototype program targeting the the products, and then track them through Hintlian, a partner in Accenture’s health pharmaceutical supply chain. Working the supply chain. According to a report and life sciences practice. with nine companies, including , released by Accenture, the project team was Accenture concluded that even though CVS Pharmacy, and Johnson & Johnson, able to read 98.6% of the case tags. When the project reached all its stated objec- Accenture tracked nearly 13,500 phar- units were inside a case, the team was able tives, “full-scale implementation on an maceutical packages using Manhattan to read 96.8% of the unit tags.” industry-wide basis will be more com- Associates’ middleware; Matrics tags, Several different scenarios were run, in- plex than many believe, requiring more readers and antennas; and Dell servers. cluding recalls and product security. Accen- time than anticipated to refine issues “Accenture worked with pharmaceutical ture also worked with the Food and Drug unique to the pharmaceutical industry.” companies to pull together a proof of Administration’s Anti-Counterfeiting Task A second series of tests will begin early concept implementation of the use of Force to learn more about how RFID can next year to gauge the business value of RFID throughout the supply chain,”says be used to snag counterfeit and gray-mar- RFID within the pharmaceutical supply Greg Gilbert, director of RFID solutions ket drugs before they get to the street. chain. at Manhattan Associates (Atlanta, GA, “We demonstrated that you can indeed –George Koroneos www.manh.com). create an RFID-enabled supply chain,

Circle/eINFO 16 Pharmaceutical Technology NOVEMBER 2003 19

IN THE FIELD

STANDARDS USP Scientific Conference Draws Crowd and Raises Questions

At the United States Pharmacopeia’s first drug manufacturer has a complete under- ate attributes. What is possible in theory Annual Scientific Meeting, held in Iselin, standing of which PCCPs are primarily re- leads to tangible manufacturing concerns. New Jersey, from 27 to 29 September, seven sponsible for the variability of critical prod- Near-infrared (NIR) and chemometrics.Ad- topic tracks engaged the more than 350 at- uct attributes, then the manufacturer can vances in NIR spectroscopy and the use of tendees in thoughtful discussion about chal- control the quality of the final product. chemometrics were discussed in the third lenges they face in formulation, quality con- Practical approaches. The second session session. Understanding how NIR behaves trol, and regulatory filings. explored PAT approaches used in other in- in real time, how chemometric data are dustries. Attendees realized they cannot an- collected, and the algorithms used to ma- Process analytical technology (PAT) swer the causality question explored in the nipulate those data can help extract new The questions of causal and correlation first session if the tools required to meas- information about the process. Because relationships, available tools, and the lim- ure critical product attributes in real or near- process-critical attributes are highly spe- its of current manufacturing systems, all real time are not fully developed. cific, a new algorithm approach has been were explored in the PAT track. Although some of these tools have been developed for transferring chemometric Causal links between processes and products. used successfully in other industries for relationships from one analyzer to another The first PAT session addressed the ques- quite some time, many cautioned that the to correct for measurement bias. tion, Can causal links be established between problems pertinent to each application are Statistical modeling. The final session, process critical control parameters (PCCPs) very specific and often significant. Also, as “Mathematics and Statistical Underpin- and critical product attributes to predict production scale increases, so do the chal- nings of Process Manufacturing,”focused product quality, and ultimately, in vivo per- lenges of implementing the on-line appli- on applying statistical modeling to PAT. formance? Attendees postulated that if a cations required to measure the appropri- USP continued on page 22

20 Pharmaceutical Technology NOVEMBER 2004 Circle/eINFO 17 www.pharmtech.com

IN THE FIELD

USP continued from page 16 review or approval process exists. The second session addressed HPLC Participants explored key issues related to For researchers trying to demonstrate column-classification systems. USP-NF component selection criteria, the false- that a new excipient is safe, speaker Robert monographs refer to general column cat- negative/false-positive quandary, the use E. Osterberg, RPh, PhD, recommended the egories, but not all brands in a category of chemometrics for diagnosing manufac- procedures outlined by Steinberg1 and Os- perform in the same way. Key questions turing problems, and outlier detection. Im- terberg2 to expedite the regulatory review were raised about how the classification ap- portant questions raised included: of a new excipient. The speaker also sug- proaches of the USP and the Product Qual- • How can you determine if an outlier re- gested that researchers continue to test ex- ity Research Institute should be made avail- sult is actually caused by model bias? cipients with active ingredients in clinical/ able to the public, and who should be • What are the regulatory implications nonclinical trials and consult the FDA and responsible for evaluating public comments of recalibrating or revalidating your CDER excipients guidances. about those systems. The discussion will model? The creation of a regulatory body to con- help USP decide on an approach for im- • Can the original assumptions made firm the safety of excipients could be an- proving the current categorization system. about variable selection or specifica- other solution, Osterberg noted. For exam- –Margareth Marques, PhD, tions be changed to modify a model ple, the industry could develop a panel such US Pharmacopoeia under new process conditions? as the Cosmetic, Toiletry, and Fragrance As- • What approaches will be allowed to sociation’s Cosmetic Ingredient Review Ex- USP approaches to microbiology achieve continuous model evaluation pert Panel, an organization that confirms In the microbiology sessions, the steriliza- and improvement? the safety of cosmetic ingredients. Accord- tion and aseptic processing discussions em- • Which statistical method is best for a ing to Osterberg, FDA already has called the phasized USP efforts to incorporate global given application? panel “an important voluntary effort.” practices. For example, the expert commit- These questions indicate that PAT still has Although excipient specifications tra- tee plans to begin a joint effort with the Eu- to overcome many hurdles before the in- ditionally have been for small molecules, ropean Pharmacopoeia to develop an in- dustry as a whole is ready to embrace it. emerging drug therapies such as bacterial/ formation chapter on moist heat validation. A broad approach to PAT, encompassing viral vectors are expanding the boundaries Experience using rapid microbiological a scientific rationale for designing and im- of what we consider to be “excipients.”As identification techniques were also dis- plementing a good manufacturing process, speaker Shireesh P. Apte, PhD, of cussed. Participants emphasized that user will have broad applicability and can give Research Inc. pointed out, such moieties requirements and expectations should be companies a regulatory advantage with could be considered excipients because defined clearly before purchasing a system. FDA.The restrictive application of PAT as they have pharmacological activity that –Thomas J. Berger, PhD, and a system of on-line measurements and may not be independent of their excipi- Ellen Tonn, Hospira, Inc. controls will likely only increase efficien- ent functionality. Thus, emerging excipi- cies for large-scale manufacturing. Such ents may require regulation based on their Making USP–NF Work for You controls will not help manufacturers that functionality and their chemistry. The four sessions of this track were designed have not taken measures to understand –Kaylynn Chiarello to educate the audience about USP activi- their processes better to establish tighter ties and processes. The first session covered specifications to control their processes. Applying chromatography standards USP fundamentals. Topics ranged from the The USP’s perspective on PAT is related During the the first chromatography ses- value of setting standards for monographs to the USP’s role. When the industry sion, questions raised about how to estab- and reference standards to the legalities of moves to innovative new technologies, the lish system-suitability led to a broader the USP–NF. The second session focused USP must incorporate them into the pub- question: How much can chromato- on how industry can effectively use the in- lic standard to continue to ensure the qual- graphic conditions be adjusted without formation in the compendia. ity of the resulting products. revalidating the analytical procedure? Par- The last two sessions focused on the prac- –Gary Ritchie, US Pharmacopeia ticipants also raised the questions of when tical applications of the USP. The dissolu- system-suitability requirements should be tion session discussed trouble-shooting ap- Regulatory paths for new excipients defined and whether USP monographs parati, proper de-aeration techniques, and A topic of discussion on day one of the “Ex- should include typical chromatograms. dissolution method development. Speak- cipients and Pharmaceutical Waters” track ers also presented the new approach USP of the USP conference is one that consis- 1 M. Steinberg et al., “A New Approach to the Safety is considering of including multiple impu- tently challenges formulators: how do you Assessment of Pharmaceutical Excipients,” Regul. rity tests in a single monograph. With this identify new excipients without fear of re- Toxicol. Pharmacol. 24 (2 pt 1), 149–154 (1996). approach, users would determine the ap- jection from the regulatory bodies? Because 2 R.E. Osterberg and N.A. See, “Toxicity of Excipi- propriate test method based on their knowl- excipients are reviewed as part of a new ents: A Food and Drug Administration Perspec- edge of the manufacturing process used. drug application, no stand-alone excipient tive,” Intl. J. Toxicol. 22 (5), 377–380 (2003). –Susan J. Schniepp, Hospira, Inc.

22 Pharmaceutical Technology NOVEMBER 2004 www.pharmtech.com

IN THE FIELD

MANUFACTURING Picking Sides: Catalytic Process Improves Drug Formulations

Distinguishing between the left- and right- program to design new catalyst candidates. handed versions of manufactured com- Testing of the initial candidates generates pounds is a long-established problem for empirical data, which is used to refine re- drug formulators. Although they are mir- sults and increase selectivity. “If the differ- ror images, the two versions frequently react entiation is just slight, we won’t see selec- differently in the body (e.g., one side being tivity,”says Marisa C. Kozlowski, PhD, Penn therapeutic and the other causing harmful associate professor of chemistry.“If it’s large, side effects). Because it’s difficult to make we will see selectivity.” A chiral catalyst developed by the University purely left- or right-handed drugs, scien- According to Kozlowski, the catalyst tech- of Pennsylvania. tists at the University of Pennsylvania (Penn, nique could offer significant advantages each method has its own applications and Philadelphia, PA, www.upenn.edu) have over other emerging chiral separation tech- advantages such as quick separation for the developed a technique that forces precur- niques. These approaches use a chromatog- chromatography approach and efficiency sors to choose a side to generate just one raphy column packed with a chiral mate- for the catalysis process. version of a drug molecule. rial to separate the compounds quickly.“The Because Penn’s technology is unpatented Penn researchers bind a single-enan- material in the column can be expensive,” and uses simple reactions and catalysts that tiomer catalyst to a precursor, which makes she points out. “In the long run, I believe most scientists can produce, their com- the two sides nonequivalent and distin- our approach will be more cost effective.” pounds can already be applied in formula- guishable. Typically, the catalyst blocks one In addition, the catalytic process is more ef- tions. At present, the group is working to ex- side of the molecule so that the reagents can ficient because half as much of the com- pand its library of known catalysts because only approach from the other side. Activa- pound is generated during the synthesis, each precursor–end-product combination tion of the coupling center is not necessary. she says.“You don’t have to discard the other requires a different catalyst. Says Kozlowski, To select a compound to bind to the sub- half of the compound after the separation,” “No one catalyst can solve every problem.” strate, the Penn team uses a computational says Kozlowski. She notes, however, that –Kaylynn Chiarello

FORMULATION FILTRATION Inconsistent Dosing Prion Removal Filter Technology from Splitting Tablets Because of the recent occurrences of vari- model for prion disease). They performed A study investigating the appropriate- ant Creutzfeldt-Jakob disease (vCJD, the an endogenous infectivity study to deter- ness of pill splitting shows that those human form of bovine spongiform en- mine the efficacy of a prototype filter in who divide a 10-mg tablet of the muscle cephalopathy, or mad cow disease) in the removing scrapie-infected prions from relaxant cyclobenzaprine hydrochloride United Kingdom, Pall Corporation (East red blood cell concentrates. After a 300- (HCl) to achieve a 5-mg dose may get Hills, NY, www.pall.com) is investigating day incubation period, none of the 20 anywhere from half to one-and-a-half the transmission and control of the blood- hamsters that received the filtered red times the intended amount of medi- borne prions (infectious protein agents) cells developed scrapie, whereas two of cine. Splitting pills may result in either thought to be responsible. the 18 hamsters that had received unfil- depriving patients of the medicine’s The researchers are developing the tered red cells developed the disease. Re- benefit or exposing them to unwanted “Leukotrap” affinity prion-reduction fil- sults showed that the filter removed in- side effects such as drowsiness.Because ter to remove both leukocytes and infec- fectious prions from red cell concentrates patients would have no guarantee of tious prions from blood before transfu- below the limit of detection of the West- consistently receiving the intended sions. Research shows the new filter has ern blot assay; a bioassay showed that the amount of medication, generic cy- an affinity for all types of prions includ- filter removed approximately 4 logs of clobenzaprine HCl 10-mg tablets ing aggregated, denatured, and normal. scrapie prions. should not be cut in half. The findings To validate the reduction of infectious The researchers plan to move the new appear in a recent issue of the Journal of prions, the researchers are studying the technology into operational trials in the American Pharmacists Association new filter using three different assays: European blood-processing centers and (www.aphanet.org). Western blot assay, bioassay, and animal hospitals in early 2005. –Megyn Bates tests for transmission of scrapie (the –Megyn Bates

24 Pharmaceutical Technology NOVEMBER 2004 www.pharmtech.com