Analysis of FDA Enforcement Reports (2012-2019) to Determine the Microbial Diversity in Contaminated Non-Sterile and Sterile Drugs

Abstract

An analysis of FDA enforcement reports from 2012 to 2019 showed that Gram-negative bacteria were the most common microbial contaminants of non-sterile drugs in the United States. Burkholderia cepacia was the number one reason for non-sterile drug recalls with 102 citations followed by Ralstonia pickettii (45 recalls) and the USP indicator, Salmonella spp. (28 recalls). Unidentified microbial contamination accounted for 77% of non-sterile and 87% of sterile drug recalls indicating extremely poor microbiology practices. The presence of yeast and mold was the reason for 52 recalls of sterile and non-sterile drugs with only 12% providing any information at the genus or species level. Gram-negative bacteria were the most common cause of microbial contamination for sterility failures with no species showing a predominant presence. However, out of specification results (34 recalls) were the most cited violation for non-sterility recalls. Most sterile drugs (1056) were recalled by the lack of sterility assurance. Undetermined cGMP issues (184 recalls) was the number one reason for lack of sterility assurance followed by compounded drugs with deficient cGMP procedures (121 recalls).

Introduction

The Food and Drug Administration (FDA) is the federal agency responsible for the regulation of pharmaceutical products in the United States of America (USA). The FDA oversees the development, manufacturing, and use of drugs by developing guidelines and regulations to guarantee safety, stability, and efficacy. Enforcement activities by the FDA range from warning letters and notice of violation letters to pharmaceutical companies. When drugs are recalled because stability, safety, or efficacy are compromised, the agency issues an enforcement report. Microbial contamination is one of the major reasons for recalls in the USA.

Analysis of FDA Enforcement Reports (2012-2019) to Determine the Microbial Diversity in Contaminated Non-Sterile and Sterile Drugs

Microbial contamination control is a fundamental requirement for sterile and non-sterile manufacturing of pharmaceutical drugs. Non-sterile pharmaceuticals are manufactured under conditions to minimize microbial contamination but the processes used during production are not monitored on a regular basis.1,2 Furthermore, the criteria for manufacturing non-sterile pharmaceuticals are completely different when compared to sterile products because of the lack of regulatory or compendial guidelines. However, according to the Code of Federal Regulations (CFR) part 211.113, companies must have appropriate written procedures, designed to prevent the presence of objectionable organisms from drug products not required to be sterile.3 This includes standard operating procedures (SOP’s) for manufacturing and quality control analysis for each nonsterile product. Written procedures for manufacturing, packaging, and quality control analysis allow reproducibility, continuity, accuracy, and process control. For instance, in sterile manufacturing, water, air, analysis, and environmental monitoring are performed on a routine basis preventing sterility failures and system breakdown. However, non-sterile manufacturing does not monitor these areas, if they monitor them at all, as frequently as sterile processes.2 Some companies performed environmental monitoring of production facilities and equipment sporadically while others did not do it on a regular basis or none at all.2 Current good manufacturing practices (cGMP) control the presence, viability, and proliferation of microorganisms. However, pharmaceutical companies follow different strategies during the manufacturing of non-sterile products. Furthermore, microbial identification of environmental isolates from non-sterile manufacturing environments varies from company to company. Some companies had a vigorous microbial identification program but others relied on simple identification schemes.2,4 Because of the infrequent and inconsistent monitoring of equipment, personnel, and environment, microbial identification from raw materials and finished products is a critical step for the quality control analysis of non-sterile pharmaceuticals. Major deviations from cGMP have been currently observed in several locations responsible for the manufacturing of non-sterile pharmaceutical products.5-7 Those deviations have not only led to major product recalls but also serious  incidents of morbidity and mortality.5,6,8

Subscribe to our e-Newsletters
Stay up to date with the latest news, articles, and events. Plus, get special offers
from American Pharmaceutical Review – all delivered right to your inbox! Sign up now!

Underwood9 clearly stated, “The microbiological quality of pharmaceutical products is influenced by the environment in which they are manufactured and by the materials used in their formulation”. If the product is not terminally sterilized, the finished product bioburden will include microbial flora from raw materials, equipment, air, personnel, and containers. Optimization of microbial contamination control requires the development and implementation of adequate systems controlling environmental conditions where product manufacturing will take place. When these systems are absent, failed, or are not validated, the manufacturer suffers by the positive evidence of objectionable microorganisms or high microbial counts in the finished product.

Jimenez10 reported the first comprehensive analysis to determine the microbial diversity in contaminated medical devices, non-sterile, and sterile drugs.10 Data collection and analysis were time consuming, repetitive, and slow. Furthermore, the information was disseminated through FDA enforcement reports, published scientific studies, and industry newsletters. Recall data showed that Burkholderia cepacia contaminated more sterile and non-sterile drugs than any other microbial species while yeast and mold contamination was found to be a major problem during manufacturing. Contamination by Gramnegative bacteria appeared to be more prevalent than with Grampositive bacteria. Large numbers of drug and medical device recalls did not provide any microbial identification at the genus or species level.10 Most medical devices and sterile drugs were recalled due to the lack of sterility assurance (LSA). Sutton and Jimenez11 provided a later update looking only at enforcement reports listed on the FDA website for the years 2004-2011 and reported similar findings regarding microbial identification, LSA, B. cepacia and mold contamination. Ever since then, the website has been significantly improved with new analytical tools to improve data mining and transparency of the different enforcement reports.

The major objective of this article is to analyze FDA enforcement reports by determining the sources of microbial contamination in nonsterile and sterile drugs from the years 2012 to 2019 and to determine if there are any changes compared to previous years. The information used for this article is listed on the FDA website regarding enforcement reports and was collected using the advance search engine:

(https://www.accessdata.fda.gov/scripts/ires/index.cfm#tabNav_advancedSearch). The recall information was restricted to enforcement reports issued from June 8, 2012 to June 21, 2019. The databases analyzed were limited to non-sterile and sterile drugs. Excel spreadsheets were used for data mining and analysis.

Bacterial Contamination of Non-Sterile Pharmaceutical Samples

An analysis of non-sterile pharmaceutical products recalled by the FDA from the years 2012-2019 demonstrated that, of the United States Pharmacopeia (USP) microbial indicators, 28 recalls were due to the presence of Salmonella spp. (Table 1). Table 1 shows data comparison with previous studies published in 2007 and 2012.10,11 Compared to what was reported for the years 2004-2011, the number of recalls by Salmonella spp. increased from two to 28.11 The numbers for 1995-2006 were also lower with only six recalls issued due to the presence of Salmonella (Table 1).10

Salmonella spp. contaminated raw materials and products such as pain relievers (Table 2). Bacterial identification was limited to the genus level with no species reported. Contamination was probably caused from raw materials that were not properly qualified with a validated microbial method to determine the microbial load or a manufacturing process that was not optimized to eliminate pathogenic microorganisms.1 Mitragyna speciosa is a tropical evergreen tree commonly known as Kratom. All contaminated samples were either Kratom raw material or formulations based upon the ingredient. Kratom is a natural product and as such, it may contain a high microbial load. The production processes for this raw material do not eliminate all microorganisms. However, an optimized manufacturing process can be designed to eliminate Salmonella from products and raw materials.1,9-11

Microbial identifi cation in recalls of raw materials, nonsterile (NS) and sterile (S) drugs (1995-2019).

Why was Salmonella spp. a reason to recall products and raw materials? The three major pharmacopoeias, USP, European (EP), and Japanese (JP) have divided non-sterile pharmaceuticals testing into two different tests: The quantitative test and qualitative test.1 The quantitative test ascertains the numbers of microorganisms, e.g., bacteria, yeast, and mold, present in a given pharmaceutical sample. The qualitative test determines the presence of specifi c pathogen indicators, e.g., Salmonella spp., Escherichia coli, Candida albicans, Clostridium spp., Staphylococcus aureus, Pseudomonas aeruginosa, and the Enterobacteriaceae family, which might cause disease to consumers or indicate the presence of other pathogenic bacteria. These indicators are representative microbial species of different types of bacterial populations. For instance, E. coli is a Gram-negative rod, capable of lactose fermentation, commonly found in fecal sources. Salmonella spp. are virulent Gram-negative rods associated to intestinal disorders while E. coli in general is not a virulent pathogen. However, some strains of E. coli are producers of toxins associated to gastrointestinal diseases. C. albicans is the most prevalent cause of fungal infections in people. Clostridium is a genus of Gram-positive anaerobic bacteria, which includes several significant human pathogens such as C. botulinum, C. difficile, C. perfringens, and C. tetani.

Pseudomonas aeruginosa is a Gram-negative non-fermentative rod, which is typically associated to opportunistic infections. S. aureus, a Gram-positive coccus, is commonly associated to skin, pneumonia, gastrointestinal, and toxic shock syndrome conditions. Some strains of S. aureus known as Methicillin-Resistant Staphylococcus aureus (MRSA) are nosocomial pathogens.

The Enterobacteriacae family, also known as bile-tolerant bacteria, comprises Gram-negative genera such as Escherichia, Salmonella, Shigella, Citrobacter, Enterobacter, Klebsiella, Proteus, etc. Most of the members of this family, other than Salmonella spp. and Shigella spp., are opportunistic pathogens widely distributed in the environment. The use of pathogen indicators does not mean that the presence of other objectionable bacteria might not be a problem during quality evaluations. However, route of application and intended use of a given product will determine if there is a risk involved when other microorganisms are present.10,11

Drug recalls for non-sterile products and raw materials based upon FDA enforcement reports (June 2012-2019) (n=163)
Drug recalls for non-sterile products and raw materials based upon FDA enforcement reports (June 2012-2019) (n=163)
Drug recalls for non-sterile products and raw materials based upon FDA enforcement reports (June 2012-2019) (n=163)

Other USP microbial indicators were cited in non-sterile recalls. C. difficile contaminated thirteen different samples (Table 1). The samples were laxative formulations and raw materials (Table 2). Contamination in raw materials such as Psyllium and lime bone gelatin promoted microbial growth in the formulations. Psyllium is a form of fiber made from the husks of the Plantago ovata plant’s seeds. The gelatin is a varied ingredient with many applications in dietary supplements and pharmaceuticals. There was no recall of non-sterile drugs by C. difficile or any other Clostridium species before 2011 (Table 1).

E. coli contaminated two samples, e.g., antacids and throat relief formulations (Table 2). E. coli contaminated one sample back in 2004-2011.11 Pseudomonas aeruginosa contaminated two samples of nasal decongestant liquids and hand sanitizer foams. Pseudomonas spp. contaminated one sample of heartburn relief medications. P. aeruginosa numbers were higher for the years 2004-2011. Nine recalls were issued due to contamination with P. aeruginosa and other Pseudomonas species (Table 1). However, in 1995-2006 enforcement reports cited P. aeruginosa nineteen times and other Pseudomonas species thirteen times (Table 1).10 One recent recall cited the presence of the Gram-negative species, Sphingomonas paucimobilis (Table 2).

Other members of the Enterobacteriacae family detected were Serratia liquefaciens with six recalls of hand sanitizer sprays without alcohol as the only products affected (Tables 1 and 2). Different species of Serratia were found in previous studies with lower recall numbers by S. fonticola and S. marcescens.11 Compared to what was reported by Jimenez10 and Sutton and Jimenez,11 Enterobacter species such as E. gergoviae and E. cloacae did not contaminate any drugs.

Staphylococcus species contaminated six samples of products and raw materials with two samples showing S. aureus, a USP indicator, and four samples with S. saprophyticus.

S. saprophyticus was always isolated with B. cepacia (Table 2). Only one recall was issue in 2004-2011 due to S. aureus contamination. A recent recall cited the presence of the Gram-positive species, Sarcina lutea (Tables 1 and 2).

B. cepacia is Still the One!

B. cepacia was again the number one microbial species contaminating non-sterile drugs. Enforcement data cited B. cepacia in 102 recalls (Table 1). B. cepacia was previously reported to be the main reason for product recalls in 1995-2006 and 2004-2011.10,11 Enforcement reports showed 30 and 35 recalls, respectively (Table 1). An average of thirteen recalls per year in 2012-2019 were calculated with more than 50% issued in the first half of 2019 (Table 3). That can be because of an increase in enforcement activities and industry’s awareness of the threat of B. cepacia to drug manufacturers leading to a more proactive approach to product testing and process control. Some of the enforcement reports contained the following statement “as a precautionary measure due to potential risk of product contamination with the bacteria B. cepacia” (Table 2). That statement does not clarify why the assessment was concluding there was a potential risk of product contamination. Based upon that statement it can be inferred that product testing might have been negative and that B. cepacia was possibly detected in water samples or during environmental monitoring. Contaminated drugs ranged from decongestants, antihemetics, nasal drops, acetaminophen syrups, antihistamines, hand sanitizers, opioids, antifungals, laxatives, ear drops, phenobarbital oral solutions, antipsychotics, laxatives, suppositories, etc. (Table 2). Other Burkholderia species contaminated several products (Table 2). Three recalls were due to the presence of B. gladioli, B. multivorans, and B. contaminans (Table 1).

Several recent morbidity and mortality incidents were found to be related to drugs contaminated with B. cepacia. An outbreak of B. cepacia complex (Bcc) pseudo bacteremia was associated to contaminated antiseptic formulations.12 B. cepacia was isolated from blood cultures of 40 patients and antiseptic formulations. The outbreak investigation determined that the formulation was misused as a skin antiseptic during blood culture. The contaminated product was discarded and the staff retrained. Another outbreak was reported at a private hospital where thirteen cancer patients undergoing chemotherapy developed B. cepacia bacteremia due to a contaminated antiemetic drug.5 Daily aseptic practices and training stopped the outbreak. Opened and unopened vials of the antiemetic drug grew B. cepacia.

Drug recalls for non-sterile products and raw materials based upon FDA enforcement reports (June 2012-2019) (n=163)
Drug recalls for non-sterile products and raw materials based upon FDA enforcement reports (June 2012-2019) (n=163)

A recent contamination of an emollient laxative triggered a multistate outbreak of B. cepacia infections in the USA.6 Five states reported more than 53 cases. Furthermore, B. cepacia was detected in the water system used for product manufacturing.7 Becker et al.8 reported a similar situation in a German hospital where two patients died of a multi-organ failure and had B. cepacia isolated from respiratory samples. The contamination was tracked to an octenidine mouthwash and the manufacturing process. Whole genome sequencing and other genotyping methods confirmed the identity of isolates from patients, products, and manufacturing samples.

Water, the most common raw material in pharmaceutical manufacturing, is also a major source of contamination and accounted for fifteen recalls of non-sterile products (Figure 1). Unfortunately, none of the recalls listed the genus or species detected only citing a high total microbial count. Water system validation is important to minimize bioburden excursions.13 B. cepacia is capable of colonizing surfaces under flowing conditions. The sanitization of the water system by heat or chemical treatment prevents microbial colonization of water lines and biofilm formation.

The persistence of B. cepacia in pharmaceutical products is due to the lack of proper cGMP and the use of compendial methods that do not provide the sensitivity and resolution to detect B. cepacia in pharmaceutical water, raw materials, and finished products.5,7,8,10,11 Cundell13 discussed the risk mitigation steps to exclude B. cepacia from non-sterile products and revealed that a proposed and timely USP test for B. cepacia was published in the September-October issue of the Pharmacopeia Forum. Some of the steps discussed by Cundell13 were:

  • Pharmaceutical ingredients selection
  • Product formulation including robust antimicrobial preservative system
  • Management of pharmaceutical water systems
  • Equipment cleaning and sanitization
  • Manufacturing processes
  • Risk-based microbial testing programs

Industrial operators and compendial methods severely underestimated B. cepacia complex (Bcc) genetic and metabolic diversity. The genome consists of more than one chromosome containing a wide variety of genes, which appeared to be acquired by horizontal transfer.14 These genes provide resistant to antibiotics, biocides, and adaptation to environmental stresses.14-17 For instance, the Bcc showed intrinsic antibiotic resistance by the release of beta-lactamases, specific efflux pumps, and the impermeability of the outer membrane to antimicrobial agents. Vigorous biofilm formation is another adaptation providing resistance to antimicrobial treatments.

Drug recalls by B. cepacia contamination from June 2012-June 2019.

B. cepacia is also capable of growing on nitro aromatic and aromatic compounds by the action of different enzymes such as monooxygenases (MO) and dioxygenases (DO). These enzymes also oxidize halogenated compounds. Nitro aromatic compounds are major components of many drugs. For instance, antipsychotic and analgesic drugs are based upon chemical structures sensitive to degradation attacks by MO and DO.

Furthermore, the Bcc is currently comprised of approximately 22 species, which are phenotypically similar.18 However, PCR based methods, pulse field gel electrophoresis (PFE), multilocus sequence typing (MLST) and whole-genome sequencing have been used to accurately identify or detect isolates from different outbreaks and contaminated products.6,7,19-21

FDA guidance has been clear and proactive regarding B. cepacia contamination.22 An article published by The Center for Drug Evaluation and Research (CDER) discussed several recall reports and inspection documents regarding points of origin for contamination and anomalies in test methods. The article also discussed the issue of objectionable microorganisms and whether B. cepacia can be included in a compendial chapter. Based upon that article, potential causes for product contamination and system breakdown were:

  • Inadequate cleaning procedures
  • Use of unsuitable grade of water (use of potable water to clean equipment)
  • Poor water system controls
  • Poor water system design
  • Inadequate testing and specifications
  • Inadequate manufacturing procedures
  • Inadequate validation or lack of environmental monitoring in critical areas

The second most common microorganism detected in non-sterile products, Ralstonia pickettii, a Gram-negative bacterial species, was always isolated with B. cepacia (Table 2). R. pickettii contaminated 45 drugs (Table 1). R. pickettii was not detected in any recalls in 2004-2011 (Table 1). However, two non-sterile drugs were recalled in 1995-2006.

Mold Contamination of Pharmaceutical Products

Reasons for non-sterile drug recalls with no microbial identification (n=548) (June 2012-2019).

Analysis of enforcement reports indicated that 52 recalls cited yeast and/or mold contamination but only six recalls listed genera or species (Table 4). Of these recalls, 27 citations were for sterile and 25 for nonsterile drugs. Yeast and/or mold recalls for non-sterile drugs were lower than in 1995-2006 when 31 citations were reported.10 However, the numbers were higher than in 2004-2011. Enforcement data from 2004-2011 showed 23 recalls due to yeast and mold contamination.11 Back in 1998-2006, mold contaminated only six sterile drugs.10

Aspergillus fumigatus and Paecilomyces saturatus simultaneously contaminated a non-sterile product while Aspergillus spp. Contaminated four different sterile formulations (Table 4). One more recall from a non-sterile drug cited Rhinocladiella similis as the contamination source. Yeast was detected in nine recalls simultaneously with B. cepacia but no identification was provided (Table 4). Yeast and mold previously isolated from contaminated products were the following: Acremonium spp., Aspergillus spp., Penicillium spp., Aspergillus sidowii, Aspergillus niger, Candida lipolytica, and Candida parapsilosis.10,11 Of all these, only Aspergillus spp. was found again in four samples of sterile products for the years 2012-2019 (Table 1).

The lack of identification of mold and yeast contamination is a worrisome trend in the pharmaceutical industry with major implications to determine proper corrective actions and process control optimization.23,24 Cundell23 previously stated the poor job performed by the pharmaceutical industry in the area of mycology. This statement is further confirmed by current recall data where only 12% of recalls showed proper identification (Table 4). Previous studies reported similar practices with only 10% and 17% of recalls showing either genus or species information.10,11 However, when sterile products were contaminated by mold, 33% showed identification at the genus level.10 Identification of yeast and mold isolates provides valuable information to determine the root cause of system failure and lack of process control.

Mycological expertise by either in-house resources or contract testing laboratories must be a basic part of the environmental and quality control program. It is inexcusable to neglect this important area of process control since molds outcompete bacteria at lower water activity environments.23 Accurate mold identification is based upon the amplification and sequencing of internal transcribed spacer (ITS) regions located between the small and large-subunits of the ribosomal gene.23 Specific PCR assays targeted different mold species in pharmaceutical products.24 Major outbreaks of fungal meningitis and endophthalmitis were associated to contamination of injectable and ocular drugs during manufacturing.25,26 Some of the mold isolates were normal flora in cleanroom environments, which evidently by the lack of process control and cGMP practices led to unfortunate events of morbidity and mortality.25-27

Unidentified Microbial Contamination of Non-Sterile Products

Reasons for sterile drug recalls due to non-sterility (n=95) (June 2012-2019).

Of 713 non-sterile recalls, 548 (77%) did not provide any information on the microorganisms responsible for the contamination. The reports did not have any genus or species level identification. This number was higher than the one reported for non-sterile drugs in 1995-2006, e.g., 43%.10 Sutton and Jimenez11 reported 28% of recalls without microbial identification in 2004-2011.11

Figure 1 shows the different reasons cited for non-sterile drug recalls. Since non-sterile drug testing looks for bioburden levels and objectionable microorganisms, it was very difficult to ascertain whether recalls were due to high levels of contamination or the presence of pathogenic microorganisms. Microbial contamination (458 recalls) was the primary reason to retrieve products from the market. The other reasons were:

  • cGMP Deviations: products manufactured by contract manufacturer under conditions that could result in possible microbial contamination (32 recalls)
  • Microbial Contamination of Non-Sterile Product (15 recalls)
  • Microbial Contamination of Non-Sterile Products: bacteria or mold/yeast counts out of specifi cation (22 recalls)
  • Out of specification total aerobic microbial count in a water sample (15 recalls)
  • Elevated counts of Gram-positive rods were found during environmental testing (3 recalls)
  • Products failed the Antimicrobial Effectiveness Test (3 recalls)
Drugs recalled by mold and yeast contamination (n=52) (June 2012-2019)
Drugs recalled by mold and yeast contamination (n=52) (June 2012-2019)

Lacking information on the microbial contaminant did not help to determine the root cause of the problem and establish proper remedial actions. Because non-sterile pharmaceuticals are allowed to have a microbial bioburden, it is critical to have a good rationale to determine what is an objectionable microorganism and the risk to consumers and manufacturing processes. Accurate microbial identification is critical to understand process control deviations and contamination excursions.28 It is indefensible not to have a proper microbial identification program with standard operating procedures (SOP’s) including analyst training. Analysts should be knowledgeable on basic microbiological procedures such as macroscopic features, e.g., looking at the plates with the naked eye to recognize whether the growth is bacterial or fungi, microscopic features, e.g., Gram-staining, spore staining, and simple biochemical tests such as catalase testing, oxidase testing, and substrate utilization tests. Ribosomal 16S rRNA gene sequencing is the current standard for bacterial identifi cation.28 The 16S ribosomal gene codes for the RNA component of the 30S subunit of the bacterial ribosome. They are essential genes which are highly conserved. Large publicly available databases containing 16S rRNA gene sequences provide enough phylogenetic information to identify bacteria at the genus and species level.

How would you know if a given microbial isolate is objectionable if you do not perform an identifi cation? The products are non-sterile and they can have a microbial load that might not be objectionable.11 Different technologies and procedures are available that rely on phenotypic and genotypic identification which provide the information necessary to determine the nature of the microorganisms present in a given sample.28,29 Sandle4 published an excellent review on the topic of microbial identification strategies in the pharmaceutical industry. However, based upon recall data some companies are still struggling to understand the risk of not having a functional microbial identification program.

Microbial Contamination of Sterile Products by Non-Sterility

Sterile drugs cited due to lack of sterility assurance (June 2012-2019) (n=1056). Only 100 recalls are shown.
Sterile drugs cited due to lack of sterility assurance (June 2012-2019) (n=1056). Only 100 recalls are shown.
Sterile drugs cited due to lack of sterility assurance (June 2012-2019) (n=1056). Only 100 recalls are shown.

When sterile drugs were recalled for non-sterility, four out of 95 recalls were due to Aspergillus species and one recall for the following bacterial species: Bacillus thuringiensis, Bacillus circulans, Variovorax paradoxus, Herbaspirillum huttiense, Achromobacter xylosoxidans, Sphingomonas paucimobilis, Klebsiella pneumoniae, and Staphylococcus warneri (Table 1, Figure 2). Of the cited eight bacterial species, three were Gram-positive and five were Gram-negative. The presence of Gram-negative bacteria in sterile products might indicate a possible problem with the water system during manufacturing while Gram-positive bacteria such as Bacillus and Staphylococcus can be the result of improper environmental control systems.4,10 The diversity of bacterial species contaminating sterile drugs in 1998-2006 was completely different with B. cepacia (five recalls), P. aeruginosa (one recall), Methylobacterium spp. (three recalls), Mycobacterium chlelonae (three recalls), Stenotrophomonas maltophilia (one recall), Ralstonia pickettii (one recall), Serratia spp. (one recall) and Bacillus licheniformis (one recall). None of these bacterial species contaminated sterile drugs in 2012-2019.

Out of specification, (OOS) results were the most cited violations with 34 followed by non-sterility (21 recalls) and mold and yeast contamination (twelve recalls) (Figure 2). Bacterial contamination, microbial contamination, failed antimicrobial effectiveness test, and container problems were the other reasons for non-sterility recalls. However, 87% of recalls did not have microbial identification. This was very similar to the numbers reported by Jimenez,10 e.g., 88%. What was the contamination source? Was it bacteria, mold, or both? Accurate investigation of OOS incidents and proper implementation of corrective actions required information that was not available from the enforcement reports. How do companies optimize microbial contamination control for sterile products if they do not identify the microorganism causing the excursion? Compendial sterility testing of pharmaceutical drugs is based upon the addition of aliquots or membranes with the concentrated samples to different types of media.30 One of the media is specific for aerobic microorganisms and the other selectively enriches for anaerobic microorganisms. The standard incubation time for both media is fourteen days. Different temperatures are used for aerobic, 25°C, and anaerobic microorganisms, 35°C. Turbidity in the enrichment media indicates positive microbial growth. The test is cumbersome and requires special gowning procedures, equipment, and laboratory facilities to reduce the risk of analyst’s and environmental contamination during testing. Sample incubation after testing requires specific incubators and daily visual readings of the enrichment cultures by the analyst to document the results. Visual readings are extremely subjective; when slow microbial growth is present, a slight pellicle can form at the bottom of the test tube or canister and will not be seen unless the sample is moderately shaken. Furthermore, in some cases when the enrichments are turbid upon the addition of samples, it is very hard to ascertain the presence or absence of microbial growth after incubation. Therefore, samples must be streaked onto solid agar media to determine the presence of any viable and culturable microorganism. This additional step extends the completion time for the test. Different technologies have been reported for rapid testing of sterile pharmaceutical products.31-34 Previous studies reported the feasibility of ATP bioluminescence and PCR analysis using universal bacterial sequences for testing of liquid products. Other studies demonstrated the evaluation of solid phase laser scanning cytometry and colorimetric sensors with reflected light to determine the amount of released carbon dioxide (CO2). Regardless of the test used, identification of the microbial contaminant will provide valuable information to complete the OOS investigation.28,35 FDA guidance to aseptic manufacturing has been quite clear stating the need to identify microorganisms when present in products and aseptic processing at the genus and species level preferably with genotypic tests.35

Lack of Sterility Assurance

Enforcement reports due to LSA (n=1056)

Based upon enforcement report data, LSA was again the number one reason for sterile drugs recalls in 2012-2019. Table 5 shows only 100 out of 1056 recalls. The drugs recalled ranged from injectable saline solutions, hormones, ophthalmic solutions, water for injection, antibiotics, vitamins, anesthetics, amino acids solutions, pain relievers, opioid analgesics, narcotics, vasopressors, etc.

Sterile drugs cited due to lack of sterility assurance (June 2012-2019) (n=1056). Only 100 recalls are shown.
Sterile drugs cited due to lack of sterility assurance (June 2012-2019) (n=1056). Only 100 recalls are shown.

Table 6 shows the different reasons for 1056 LSA citations. The number one reason cited was undetermined cGMP issues (184 recalls) followed by compounded drugs with deficient cGMP procedures (121 recalls). Compounded drugs accounted for a large number of deviations. Out of 1056 citations, 523 (50%) were tracked to compounded products. Friedman36 stated that three prevalent themes central to aseptic processing contamination are poor personnel practices, loss of environmental control, and flawed operational design. Looking at specific reasons in Table 6, most problems were related to faulty quality control procedures, unreliable sterility and stability data, lack of environmental monitoring, media fill failures, and possible yeast and mold contamination.

LSA was also the number one reason for sterile drug recalls in previous years.10,11 Jimenez reported that 75% (71 out of 95) of sterile drugs were recalled by LSA.10 Therefore, the numbers found in 2012-2019 were 10 times higher than previously reported. The fact that it was determined that the probability and risk of introducing microorganisms into products were beyond acceptable levels indicated the complete lack of process design and control.36,37

Sterilization is a process that removes and kills all microorganisms through a chemical agent or physical process.30 However, there is no absolute certainty that all the units will be sterile. This is because not all units are tested for sterility. To provide that kind of degree of assurance, all units must be shown to be sterile. This cannot be accomplished unless all units are destroyed. Therefore, the sterility of a pharmaceutical lot is described as a probability where the likelihood of a contaminated unit or article is acceptably remote. Such a state of Sterility Assurance Levels (SAL) can only be established with adequate validated sterilization cycles and aseptic processing under appropriate cGMP. Furthermore, environmental monitoring of facilities, personnel, and processes is a major component during process control of sterile manufacturing and testing.

The likelihood of a product to be sterile is best explained in terms of the probability of microorganisms to survive the treatment process. For pharmaceutical sterilization procedures, the standard probability is less than one in one million units processed (<10-6).38 For instance, for a product containing 103 spores, an inactivation factor of 10-9 will be needed to give a sterility assurance level of 10-6. This indicates that there is a probability of less than one in a million of microbial survivors to be present in a given sterile batch. Therefore, the sterilization process will need to produce a lethality level that will kill all microorganisms.

Some of the most common procedures recommended to sterilize a product are:

  • Filtration
  • Steam sterilization
  • Dry heat sterilization
  • Ionizing irradiation
  • Ethylene oxide
Enforcement reports due to LSA (n=1056)

The choice depends on the capacity of the formulation and package to resist any of the above treatments. However, most of the drugs shown in Tables 5 and 6 were manufactured by aseptic processing.5,30,37 Questionable package integrity increased the probability of microorganisms to compromise drug safety and potency (Table 6, reason 17). Deficient sterilization validation allowed the possibility of microbial survival and growth contaminating the finished drug product (Table 6, reason 10). Unsuccessful media fill operations resulting in contaminated vials indicated possible flaws in aseptic process design and execution (Table 6, reason 27). Improperly designed environmental monitoring program increased the risk of introducing microorganisms into the process resulting in major production losses (Table 6, reason 11). Unaudited contract-testing laboratory protocols and procedures endangered test reliability and compliance (Table 6, reasons 15 and 23). Non-validated processes to reduce and eliminate endotoxins from manufacturing resulted in OOS for endotoxins (Table 6, reason 18). Unreliable sterility and stability data compromised product’s safety and efficacy (Table 6, reason 7). Improperly designed and nonvalidated processes increased the chance of microbial survival and contamination (Table 6, reasons 13, 23, 24). Deficient documentation and improper investigation procedures compromised data integrity and reliability (Table 6, reasons 11, 18). Improper facility and process design enhanced the risk of introducing microorganisms to processes and products (Table 6, reasons 30, 32). Poor aseptic practices during manufacturing compromised the sterility of products intended to be sterile (Table 6, reasons 16, 29).

Conclusions

Based upon publicly available FDA enforcement reports from June 2002-2019, microbial contamination control continues to be a major challenge for the pharmaceutical industry and regulatory agencies. Recalls due to B. cepacia and other bacterial species, mold, and LSA were higher than in previous years. Significant increases were also observed related to unidentified microbial contamination of nonsterile products. Enforcement activities appeared to be higher than in previous years specially when related to compounded drugs.

The FDA website provided a comprehensive search tool to analyze enforcement reports of non-sterile and sterile drugs. The remarkable improvements performed to the FDA website through the years facilitated and optimized data mining, navigation, and search of a large number of enforcement reports. These improvements provided the discovery of patterns and trends to organize information into a comprehensible structure for further analysis. The author is extremely grateful for the diligence of the FDA in providing a transparent and clear resource to ascertain the microbiological analysis of drug recalls.

The development and manufacturing of sustainable, safe, efficacious, and stable pharmaceutical products are major goals of the pharmaceutical industry. FDA enforcement data provided valuable information to determine the major sources of microbial contamination in drugs and current non-compliance practices in industry that led to major production losses, product recalls, and unfortunate incidents of morbidity and mortality. Learning from the information in the enforcement reports will provide effective and proactive approaches to eliminate improper practices leading to the optimization of process control and manufacturing of pharmaceutical products. The information also conveys the agency’s current thinking on compliance and enforcement priorities.

References

  1. Jimenez, L. Microbial limits. In Microbial Contamination Control in the Pharmaceutical Industry. Luis Jimenez (Editor). Marcel Dekker Publishers, New York, N.Y., Chapter 2, 15-44, (2004).
  2. Jimenez, L. Environmental monitoring. In Microbial Contamination Control in thePharmaceutical Industry. Luis Jimenez (Editor). Marcel Dekker Publishers, New York, N.Y., Chapter 5, 103-132, (2004).
  3. Federal Register of the United States of America. Code of Federal Regulations. Subpart F: Production and Process Control, 21 Part 211.113: Control of microbiological contamination, Rockville, Maryland, (2018).
  4. Sandle, T. Microbial identification strategy for pharmaceutical microbiology. Journal of GXP Compliance 21, 11-20, (2017).
  5. Singhal, T., Shah, S., Naik, R. Outbreak of Burkholderia cepacia complex bacteremia in a chemotherapy day care unit due to intrinsic contamination of an antiemetic drug. Indian Journal of Medical Microbiology 33, 117-119, (2015).
  6. Centers for Disease Control and Prevention. Multistate outbreak of Burkholderia cepacia bloodstream infections associated with contaminated prefilled saline flush syringes. https://www.cdc.gov/hai/outbreaks/b-cepacia-saline-flush/index.html, (2016).
  7. Marquez, L., Jones, K.N., Whatley, E.M., Koy, T.H., Revell, P.A., Taylor, R.S., Brooke Bernhardt, M., Wagner, J.L., Dunn, J.J., LiPuma, J.J., Cambell, J.R. An outbreak of Burkholderia cepacia complex infections associated with contaminated liquid docusate. Infection Control & Hospital Epidemiology 38, 567-573, (2017).
  8. Becker, S.L., Berger, F.K., Feldner, S.K., Karilova, I., Haber, M., Mellmann, A., Schafers, H.J., Gartner, B. Outbreak of Burkholderia cepacia complex infections associated with contaminated octenidine mouthwash solution, Germany, August to September 2019. European Surveillance 23 (42):pii=1800540. https://doi.org/10.2807/1560-7917. ES.2018.23.42.1800540, (2018).
  9. Underwood, E. Ecology of microorganisms as its affects the pharmaceutical industry. In W.B. Hugo and A.B. Russell eds. Pharmaceutical Microbiology, 6th Edition, Blackwell Science, Oxford, England, Chapter 17, 339-354, (1998).
  10. Jimenez, L. Microbial diversity in pharmaceutical product recalls and environments. PDA Journal of Pharmaceutical Science and Technology 61, 383-399, (2007).
  11. Sutton, S.W., and L. Jimenez. A review of reported recalls involving microbiological control 2004-2011 with emphasis on FDA considerations of “Objectionable Organisms”. American Pharmaceutical Review 15, 42-57, (2012).
  12. Ko, S., An, H.S, Bang, J.S., Park, S.W. An outbreak of Burkholderia cepacia complex pseudobacteremia associated with intrinsically contaminated commercial 0.5% chlorhexidine solution. American Journal of Infection Control 43, 266-268, (2015).
  13. Cundell, A. Excluding Burkholderia cepacia complex from aqueous, non-sterile drug products. https://www.americanpharmaceuticalreview.com/Featured-Articles/358427-Excluding-i-Burkholderia-cepacia-i-complex-from-Aqueous-Non-Sterile-Drug-Products/, posted on February 12, (2019).
  14. Rodley, P.D., Romling, U., Tummler, B. A physical genome map of the Burkholderia cepacia type strain. Molecular Microbiology 17, 57-67, (1995).
  15. Juhas, M., Stark, M., Von Mering, C., Lumjiaktase, P., Crook, D.W., Valvano, M.A., Eberl, L. High confidence prediction of essential genes in Burkholderia cenocepacia. PLoS ONE 7(6): e40064. doi:10.1371/journal.pone.0040064, (2012).
  16. Lessie, T.G., Hendrickson, W., Manning, B.D., Devereaux, R. Genomic complexity and plasticity of Burkholderia cepacia. FEMS Microbiology Letters 144, 117-128, (1996).
  17. Rushton, L., Sass, A., Baldwin, A., Dowson, C.G., Donoghue, D., Mahenthiralingam, E. Key role for efflux in the preservative susceptibility and adaptive resistance of Burkholderia cepacia complex bacteria. Antimicrobial Agents and Chemotherapy 57, 2972-2980, (2013).
  18. Devanga, N.K., Veeraraghavan, B. Accurate identification and epidemiological characterization of Burkholderia cepacia complex: an update. Ann. Clin. Microbiol. Antimicrob. Feb 4. doi: 10.1186/s12941-019-0306-0, (2019).
  19. Jimenez, L., Jashari, T., Vasquez, J., Zapata, S., Bochis, J., Kulko, M., Ellman, V., Gardner, M., Choe, T. Real-Time PCR detection of Burkholderia cepacia in pharmaceutical products contaminated with low levels of bacterial contamination. PDA Journal of Pharmaceutical Science and Technology 72:73-80, (2018).
  20. Jimenez, L., Smalls, S. Molecular detection of Burkholderia cepacia in toiletry, cosmetic and pharmaceutical raw materials and finished products. Journal of AOAC International 83, 963-966, (2000).
  21. Attia, M.A., Ali, A.E., Essam, T.M., Amin, M.A. Direct detection of Burkholderia cepacia in susceptible pharmaceutical products using semi-nested PCR. PDA Journal of Pharmaceutical Science and Technology 70, 99-108, (2016).
  22. Torbeck, L., Raccasi, D., Guilfoyle, D. E., Friedman, R. L., Hussong, D. Burkholderia cepacia: This decision is overdue. PDA Journal of Pharmaceutical Science and Technology 65, 535–543, (2011).
  23. Cundell, A.M. Mold monitoring and control in pharmaceutical manufacturing areas. https://www.americanpharmaceuticalreview.com/Featured-Articles/190686-Mold-Monitoringand-Control-in-Pharmaceutical-Manufacturing-Areas/, posted July 30, (2016).
  24. Jimenez, L., Smalls S., Ignar R. Use of PCR analysis for detecting low levels of bacteria and mold contamination in pharmaceutical samples. Journal of Microbiological Methods. 41, 259-65, (2000).
  25. Ahearn, D.G., Stulting, R.D. Moulds associated with contaminated ocular and injectable drugs: FDA recalls, epidemiology considerations, drug shortages, and aseptic processing. Medical Mycology 56, 389-394, (2018).
  26. Smith, R.M., Schaefer, M.K., Kainer, M.A., Wise, M., Finks, J., Duwve, J., Fontaine, E., Chu, A., Carothers, B., Reily, A., Fiedler, J., Wiese, A.D., et al. Fungal infections associated with contaminated methylprednisolone injections. New England Journal of Medicine 369, 1598-1609, DOI: 10.1056/NEJMoa1213978, (2013).
  27. Sandle, T. Fungal contamination of pharmaceutical products: the growing menace. European Pharmaceutical Review 19, 68-71, (2014).
  28. Cundell, A.M. Microbial identification strategies in the pharmaceutical industry. PDA Journal of Pharmaceutical Science and Technology 60, 111-23, (2006).
  29. Jimenez, L. Molecular applications to pharmaceutical processes and clean room environments. PDA Journal of Pharmaceutical Science and Technology 65, 242-253, (2011).
  30. Jimenez, L. Sterility testing and sterility procedures. In Microbial Contamination Control in the Pharmaceutical Industry. Luis Jimenez (Editor). Marcel Dekker Publishers, New York, N.Y., Chapter 4, pages 77-102, (2004).
  31. Jimenez, L., Ignar, R., D’Aiello, R., Grech, P. Use of PCR analysis for sterility testing in pharmaceutical environments. Journal of Rapid Methods and Automation in Microbiology 8, 11-20, (2000).
  32. Gray, J.C., Morandell, D., Gapp, G., Le Goff, N., Neuhaus, G., Staerk, A. Identification of microorganisms after Milliflex rapid detection-A possibility to identify nonsterile findings in the Milliflex rapid sterility test. PDA Journal of Pharmaceutical Science and Technology 65, 42-54, (2011).
  33. Smith, R., Von Tress, M., Tubb, C., Vanhaecke, E. Evaluation of the ScanRDI as a rapid alternative to the pharmacopoeial sterility test method: comparison of the limits of detection. PDA J. Pharm. Sci. Technol. 64, 356-363, (2010).
  34. Jimenez, L., Rana, N., Amaraj, J., Walker, K., Travers, K. Validation of the BacT/ALERT® 3D system for rapid sterility testing of biopharmaceutical samples. PDA J. Pharm. Sci. and Technol. 66, 38-54, (2012).
  35. U.S. Food and Drug Administration. Guidance on sterile drug products produced by aseptic processing. https://www.fda.gov/regulatory-information/search-fda-guidancedocuments/sterile-drug-products-produced-aseptic-processing-current-goodmanufacturing-practice, September, (2004).
  36. Lolas, A.G., Metcalfe, J. Evaluation of microbial growth potential of pharmaceutical drug products and quality by design. PDA Journal of Pharmaceutical Science and Technology 65, 63-70, (2011).
  37. Friedman, R.L. Aseptic processing contamination case studies and the pharmaceutical quality system. PDA Journal of Pharmaceutical Science and Technology 59, 118-126, (2005).
  38. Woedtke, T., Kramer, A. The limits of sterility assurance. GMS Krankenhhyg Interdiszip. 3(3): Doc19. Published online 2008 Sep 3. PMCID: PMC2831250, (2008)
  • <<
  • >>
  • Count-Cal Count Precision Size Standard Particles

    The Thermo Scientific* COUNT-CAL Count Precision Standards are easy-to-use liquid particle count controls. COUNT-CAL* is a cost effective, convenient product for validating liquid particle counters. Packaged in single-use bottles and intended to be sampled directly from the bottle, it eliminates the need for serial dilutions and extensive sample handling, thereby minimizing contamination. It was developed for manufacturers of parenteral drugs and ophthalmic solutions seeking interim verification of USP 788 and 789 (Particulate Matter in Injections and Particulate Matter in Ophthalmic Solutions).

Join the Discussion