FOOD SAFETY and FOOD SECURITY Voeller V05-Fﬁrs.Tex V1 - 12/04/2013 4:34Pm Page Ii Voeller V05-Fﬁrs.Tex V1 - 12/04/2013 4:34Pm Page Iii

Voeller V05-fﬁrs.tex V1 - 12/04/2013 4:34pm Page ii Voeller V05-fﬁrs.tex V1 - 12/04/2013 4:34pm Page i

FOOD SAFETY AND FOOD SECURITY Voeller V05-fﬁrs.tex V1 - 12/04/2013 4:34pm Page ii Voeller V05-fﬁrs.tex V1 - 12/04/2013 4:34pm Page iii

FOOD SAFETY AND FOOD SECURITY

Edited by JOHN G. VOELLER Black & Veatch Voeller V05-fﬁrs.tex V1 - 12/04/2013 4:34pm Page iv

Published by John Wiley & Sons, Inc., Hoboken, New Jersey Published simultaneously in Canada

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at http://www.wiley.com/go/permission.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002.

Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com. ePDF: 9781118651810 ePub: 9781118651827

Printed in the United States of America

10987654321 Voeller v05-ftoc.tex V1 - 12/05/2013 7:55pm Page v

CONTENTS

Preface vii

1. Microbiological Detectors for Food Safety Applications 1 2. Processing and Packaging that Protects the Food Supply Against Intentional Contamination 27 3. Early Detection and Diagnosis of High-Consequence Plant Pests in the United States 43 4. The Role of Food Safety in Food Security 61 5. Decontamination and Disposal of Contaminated Foods 69 6. Pulsenet: A Program to Detect and Track Food Contamination Events 83 7. Insects as Vectors of Foodborne Pathogens 97 8. Farm Level Control of Foreign Animal Disease and Food-borne Pathogens 111 9. Potential for Human Illness from Animal Transmission of Food-borne Pathogens 133 10. Mitigating Consequences of Pathogen Inoculation into Processed Food 149

Index 157

v Voeller v05-ftoc.tex V1 - 12/05/2013 7:55pm Page vi PREFACE

Adapted from the Wiley Handbook of Science and Technology for Homeland Security. The topic of homeland security did not begin with the World Trade Center or the IRA or the dissidents of past empires, but began when the concept of a nation versus a tribe took root and allegiance to a people was a choice, not a mandate. The concept of terrorism is part of homeland security but there are other risks to homeland security; such as those that come from Mother Nature or negligence of infrastructure maintenance. Indeed, these factors have much higher probabilities of causing substantial damage and loss of life than any group of terrorists could ever conceive. Hence, the focus here is on situations that put humans at risk and can disrupt and damage infrastructure, businesses, and the environment, and on scientific and technological tools that can assist in detecting, preventing, mitigating, recovering, and repairing the effects of such situations. The number of science and technology (S&T) related topics that are involved in the physical, cyber and social areas of homeland security includes thousands of specialties in hundreds of disciplines so no single collection could hope to cover even a majority of these. Instead, our intention is to discuss selected topics in ways that will allow readers to acquire basic knowledge and awareness and encourage them to continue developing their understanding of the subjects. Naturally, in the context of homeland security and counterterrorism, some work has to be classified so as not to “communicate our punches” to our adversaries and this is especially true in a military setting. However, homeland security is concerned with solutions to domestic situations and these must be communicated to officials, law enforcement, and the public. Moreover, having experts speak in an open channel is important for inform- ing researchers, academics, and students so that they can work together and increase our collective knowledge. There are many ways to address homeland security concerns and needs, and many different disciplines and specialties. An ongoing open conversation among experts which will allow them to connect with others and promote collaboration, shared learning and new relationships is needed. Certainly, creating a forum in which theories, approaches, vii viii PREFACE solutions and implications could be discussed and compared would be beneficial. In addition, reliable sources from which experts and lay persons alike could learn about various facets of homeland security are needed. It is equally important that policy and decision makers get the full picture of how much has been done and how much still needs to be done in related areas. Even in places that have dealt with terrorism for over a century, there are no strong, cost-effective solutions to some of the most pressing problems. For example, from a distance, we have very limited ability to spot a bomb in a car moving toward a building to allow decision making on whether to destroy or divert the car before it can damage the target. Even simpler, the ability to spot a personnel-borne improvised explosive device (IED) in a crowd coming into a busy venue is still beyond our collective capability. Therefore, the bounding of what we know and don’t know needs to be documented. Finding additional uses for technologies developed originally to solve a homeland security problem is one of the most important aspects of the economics involved. An inescapable issue in many areas of homeland security S&T, is that even a successful solution when applied to only a small market will likely fail because of insufficient returns. For example, building a few hundred detectors for specific pathogens is likely to fail because of limited demand, or it may never even receive funding in the first place. The solution to this issue is finding multiple uses for such devices. In such a case, a chemical detector for contraband or dangerous materials could be used also to detect specific air pollutants in a building; thus, help allergy sufferers. In this way capabilities developed for homeland security may benefit other, more frequently needed uses, thereby making the invention more viable. The editors of this work have done a superb job of assembling authors and topics and ensuring good balance between fundamentals and details in the chapters. The authors were asked to contribute material that was instructional, discusses a specific threat and a solution, or provides a case study on different ways a problem could be addressed and what was found to be effective. We wanted new material where possible. The authors have produced valuable content and worked hard to enhance quality and clarity of the chapters. And finally, the Wiley staff has taken on the management of contributors with patience and energy beyond measure.

Senior Editor John G. Voeller 1 MICROBIOLOGICAL DETECTORS FOR FOOD SAFETY APPLICATIONS

Evangelyn C. Alocilja and Sudeshna Pal Biosystems and Agricultural Engineering, Michigan State University, East Lansing, Michigan

1.1 BIOSECURITY AND FOOD SAFETY THREATS

The complexity of the US food supply chain from cradle to grave provides numerous entry points and routes in which (inadvertent and intentional) contaminants and pathogens can be introduced into the nation’s food system. For example, a simple hamburger, consisting of a bun, a beef patty, tomato, lettuce, cheese, and onion, is made of at least 50 ingredients which could include hundreds of sources when we consider the raw materials, processing, transportation, and finished product. Furthermore, these ingredients may come from across the globe, crossing the US border in less than 24 h. The recent scandal on melamine-tainted pet foods (and maybe human food through melamine-tainted animal feed) is one example of how the food supply can potentially be sabotaged. The use of microorganisms as biological weapons has long been reported in history. One of the first major attacks that have been reported occurred in the 14th century with Yersenia pestis during the siege of Kaffa [1]. The most recent was the deliberate release of Bacillus anthracis spores through the postal system in the United States in October 2001, shortly after the terrorist attack, resulting in 22 cases of anthrax and five deaths [2]. Inhala- tional anthrax has a high mortality rate of about 100% and the spore forms of the bacteria are very stable under harsh environmental conditions. The Centers for Disease Con- trol and Prevention (CDC, http://www.bt.cdc.gov/agent/agentlist.asp) and the National Institute of Allergy and Infectious Diseases (NIAID, http://www3.niaid.nih.gov/topics /BiodefenseRelated/Biodefense/ research/CatA.htm) have classified B. anthracis as a Biodefense Category A agent because it can be easily transmitted from person to person, can cause high mortality with potential for major public health impact, may cause public panic and social disruption, and requires special action for public health preparedness. It is estimated that the release of 50 kg of dried anthrax spores for 2 h can lead to a complete breakdown in medical resources and civilian infrastructure in a city of 500,000 inhabitants [3]. B. anthracis is a gram-positive, nonmotile, facultatively anaerobic, spore-forming, rod-shaped bacterium and is the etiological agent of anthrax. Anthrax is primarily a

1 2 MICROBIOLOGICAL DETECTORS FOR FOOD SAFETY APPLICATIONS

zoonotic disease but all mammals, particularly humans, are prone to this disease. The spore forms of B. anthracis are highly resistant to adverse environmental conditions, such as heat, ultraviolet and ionizing radiation, pressure, and chemical agents. They are able to survive for long periods of time in contaminated soils and this account for the ecological cycle of the microorganism. The vegetative cells of the bacterium are square-ended and capsulated having a size range of 3 to 5 μm while the spores are elliptical with a size range of 1 to 2 μm[4]. The primary virulence factors of B. anthracis are toxin production and capsule formation. Virulent strains of the microorganism carry two large plasmids pXO1 and pXO2 which encode these virulence factors. The plasmid pXO1 carries the structural genes for the anthrax toxin proteins pagA (protective antigen), lef (lethal factor), and ef (edema factor); two trans-acting regulatory genes atxA and pagR; a gene encoding type I topoiso- merase, topA; and a three gene operon, gerX , which affects germination. Plasmid pXO2 carries three genes which encode capsule synthesis: capA, capB, and capC ; a gene associated with capsule degradation, dep; and a trans-acting regulatory gene acpA [5]. None of the three toxin proteins are toxic separately. Toxicity is associated with the formation of binary exotoxins. The association of pagA and lef results in the formation of lethal toxin (LTx), which provokes lethal shock in animals, while the association of pagA and ef forms the edema toxin (ETx), which produces edema in the skin [6]. B. anthracis spores can enter the human host through the skin (cutaneous route), ingestion (gastrointestinal route), and inhalation (pulmonary route). Ingesting food products contaminated with the spores can lead to gastrointestinal anthrax. In this manner, anthrax spores may cause lesions from the oral cavity to the cecum [7]. Cases of gastrointestinal anthrax have been reported through ingesting undercooked meat from animals [8]. The disease is characterized by fever, nausea, vomiting, abdominal pain, and bloody diarrhea [8]. Gastrointestinal anthrax has been reported to cause fatalities in 25-60% of cases (CDC, 2001). In some community-based studies, cases of gastrointestinal anthrax outnumbered those of cutaneous anthrax [7]. Awareness of gastrointestinal anthrax in a differential diagnosis remains important in anthrax-endemic areas but now also in settings of possible bioterrorism. The inhalational form of anthrax is considered the most dangerous among the three routes, having a mortality rate close to 100% (CDC, 2001). The inhaled spores reach the alveolus where they are phagocytosed by macrophages and transported to the medi- astinal lymph nodes, where spore germination can occur in up to 60 days. Following germination, the disease progresses rapidly resulting in the production of exotoxins that cause edema, necrosis, and hemorrhage [4]. Diagnosis is difﬁcult in both gastrointestinal and inhalational forms, resulting in the disease rapidly becoming treatment-resistant and fatal. In addition to intentional contaminations, we have recently faced unintentional food poisoning through pathogen-tainted products which caused recalls on these products. In September 2007, a major meat processing company recalled up to 9,843 mt (21.7 million lb) of ground beef due E. coli O157:H7 contamination; it was one of the largest meat recalls in US history. This contamination sickened 30 people in eight states. On October 5, 2007, that company announced that it was closing its business.1 Contamination of meat products by foodborne pathogens is increasingly a major food safety and economic concern. Billions of dollars are lost every year in medical costs, productivity, product

1http://www.msnbc.msn.com/id/21149977/ BIOSECURITY AND FOOD SAFETY THREATS 3 recalls, and jobs as a result of pathogen-contamination outbreaks. In the United States, there are up to 33 million cases of human illness each year from microbial pathogens in the food supply with an associated cost of $2–4 billion in 2006.2 NIAID has identified the following microbes as foodborne and waterborne pathogens: diarrheagenic Escherichia coli, Salmonella species, pathogenic Vibrio s, Shigella species, Listeria monocytogenes, Campylobacter jejuni, Yersinia enterocolitica, caliciviruses, Hepatitis A, Cryptosporidium parvum, Cyclospora cayatanensis, Giardia lamblia, Entamoeba histolytica, Toxoplasma, and Microsporidia. These organisms are classified as Category B because they are moderately easy to disseminate, result in moderate morbidity rates, and require specific enhancements of CDC’s diagnostic capacity and enhanced disease surveillance (http://www.bt.cdc.gov/agent/agentlist.asp). In general, the causes of foodborne illness include viruses, bacteria, parasites, fungi, toxins, and metals with the symptoms ranging from mild gastroenteritis to life-threatening neurological, hepatic, and renal problems. It is estimated that foodborne diseases cause approximately 76 million illnesses, including 325,000 hospitalizations and 5000 deaths in the United States each year [9]. Of these, known pathogens account for an estimated 14 million illnesses, 60,000 hospitalizations, and 1800 deaths indicating that these pathogens are a substantial source of infectious diseases [9]. Researchers at the Economic Research Service (ERS) of the US Department of Agriculture (USDA) estimate that the total annual medical cost associated with foodborne illness caused by pathogens is $6.5–9.4 billion. Recent foodborne disease outbreaks involved E. coli O157:H7 in spinach in 2007, and cookie dough in June 2009, and Salmonella in peanut butter in January 2009. E. coli are bacteria that naturally occur in the intestinal tracts of humans and warm-blooded animals to help the body synthesize vitamins. A particularly dangerous type is the enterohemorrhagic E. coli O157:H7 or EHEC. In 2000, EHEC was the etiological agent in 69 confirmed outbreaks (twice the number in 1999) involving 1564 people in 26 states [10]. Of the known transmission routes, 69% were attributed to food sources, 11% to animal contact, 11% to water exposures, and 8% to person-to-person transmission [10]. E. coli O157:H7 produces toxins that damage the lining of the intestine, cause anemia, stomach cramps, and bloody diarrhea, and a serious complication called hemolytic uremic syndrome (HUS) and thrombotic thrombocytopenic purpura (TTP) [11]. In North America, HUS is the most common cause of acute kidney failure in children, who are particularly susceptible to this complication. TTP has a mortality rate of as high as 50% among the elderly [12]. Recent food safety data indicates that cases of E. coli O157:H7 are rising in both the United States and other industrialized nations [13]. Human infections with E. coli O157:H7 have been traced back to individuals having direct contact with food in situations involving food handling or food preparation. The most recent E. coli O257:H7 outbreak covering 29 states involved eating raw refrigerated prepackaged cookie dough [14]. In addition to human contamination, E. coli O157:H7 may be introduced into food through meat grinders, knives, cutting blocks, and storage containers. E. coli O157:H7 has also been found in drinking water that has been contaminated by runoff from livestock farms as a result of heavy rains. Regardless of source, E. coli O157:H7 has been traced to a number of food products including meat and meat products, apple juice or cider, milk, alfalfa sprouts, unpasteurized fruit juices, dry-cured salami, lettuce, game meat, and cheese curds [11, 15]. Possible points of entry into the

2http://www.ers.usda.gov/Data/FoodborneIllness/ 4 MICROBIOLOGICAL DETECTORS FOR FOOD SAFETY APPLICATIONS food supply chain include naturally occurring sources from wild animals and ecosystems, infected livestock, contaminated processing operations, and unsanitary food preparation practices. Salmonella enterica serovar Typhimurium and Salmonella enterica serovar Enteritidis are the most common Salmonella serotypes found in the United States. According to CDC, salmonellosis is the most common foodborne illness [16]. Over 40,000 actual cases are reported yearly in the U.S. [17]. Approximately 500 [9] to 1,000 [18] persons die annually from Salmonella infections in the United States. The estimated annual cost of human illness caused by Salmonella is $3 billion [9]. Salmonella Enteritidis has frequently been observed as a contaminant in foods such as fresh produce, eggs, and poultry products. While various Salmonella species have been isolated from the outside of egg shells, presence of Salmonella Enteritidis inside the egg is of great concern as it suggests vertical transmission, that is, deposition of the organism in the yolk by an infected hen (prior to shell deposition) [19]. The recent outbreak of Salmonella involving peanut butter in January 2009 hit almost every state in the United States. Human Salmonella infection can lead to enteric (typhoid) fever, enterocolitis, and systemic infections by non-typhoid microorganisms. Typhoid and paratyphoid strains are well-adapted for invasion and survival within host tissues, causing enteric fever which is a serious human disease. Non-typhoid Salmonella causes salmonellosis, which is manifested as gastroenteritis with diarrhea, fever, and abdominal cramps. Severe infection could lead to septicemia, urinary tract infection, and even death in at-risk populations (young, elderly, and immunocompromised individuals). Raw meats, poultry, eggs, milk and dairy products, ﬁsh, shrimp, frog legs, yeast, coconut, sauces and salad dressing, cake mixes, cream-ﬁlled desserts and toppings, dried gelatin, peanut butter, cocoa, and chocolate are some of the foods associated with Salmonella infection.

1.2 DETECTION

The detection and identification of these foodborne pathogens in raw food materials, ready-to-eat food products, restaurants, processing and assembly lines, hospitals, ports of entry, and drinking water supplies continue to rely on conventional culturing techniques. Conventional methods involve pre-enrichment, selective isolation, and biochemical screening, as well as serological confirmation for certain pathogens. Hence, a complex series of tests is often required before any identification can be confirmed. These methods are laborious and may require a certain level of expertise to perform. Though these methods are highly sensitive and specific, they are elaborate, laborious, and typically require 2–7 days to obtain conclusive results [15]. Their results are not available on the time-scale desired in the food quality assurance or clinical laboratory, which has safety, cost, and quality implications for the food, medical, and biodefense sectors. Rapid detection methods for pathogens have hence become a necessity. Currently, the three most popular methods for detecting pathogens are: microbial culturing followed by biochemical identification, enzyme-linked immunosorbent assay (ELISA), and polymerase chain reaction (PCR) assay. Conventional microbial culturing techniques are very sensitive; however, they include multiple steps in the assay and require pre-enrichment steps and time consuming processes. For example, conventional detection and specific identification of B. anthracis require complex techniques and laborious methods because of the genetic similarities among various Bacillus species as well DETECTION 5 as their existence in both spore forms and vegetative state. B. anthracis is identified using standard biochemical techniques, such as its sensitivity to penicillin, nonmotility, non β-hemolytic behavior on sheep or horse blood agar plates, and its susceptibility to lysis by gamma phage. It has been reported that identification of B. anthracis by initial blood culturing requires 6–24 h for growth, which is followed by morphological and biochemical identification that requires an additional 12–24 h, and finally, definitive identification that requires an additional 1–2 days [20]. B. anthracis is also shown to selectively grow on polymyxin-lysozyme EDTA-thallous acetate (PLET) agar which requires 1–2 days for growth followed by further confirmation [21]. ELISA is a diagnostic tool to detect the presence of antibody-antigen reaction in a sample. An unknown amount of antigen is affixed to a surface, and then a specific antibody is washed over the surface so that it can bind to the antigen. This antibody is linked to an enzyme, and in the final step a substance is added that the enzyme can convert to some detectable signal. ELISA is becoming very popular for food safety monitoring. PCR is gaining popularity in non-culture-based detection schemes. It is highly sensitive and able to detect the presence of just one cell. However, PCR technology has some disadvantages such as the requirement of expensive equipment, skilled personnel to perform assays, DNA extraction stages which increase the detection time, and prior information of target DNA sequences. Biosensors can play a role in the rapid test market. Biosensor technology is emerging as a promising field for rapid detection of microbial pathogens. A biosensor is an analytical device that integrates a biological sensing element with an electrical transducer to quantify a biological event (e.g. an antigen-antibody reaction) into an electrical output. The basic concept of operation of a biosensor is illustrated in Figure 1.1. The biological sensing element may include enzymes, antibodies, DNA probes, aptamers, molecularly imprinted polymers, and whole cells. Depending on the transducing mechanism, biosensors can be electrochemical, electrical, optical, mechanical, and magnetic. They can be operated in a reagent-less process enabling the creation of user-friendly and field-ready devices. Some of the major attributes of biosensor technology are its specificity, sensitivity, reliability, portability, real-time analysis, and simplicity of operation. Biosensors are needed to quickly detect disease-causing agents in food, in order to ensure continued safety of the nation’s food supply. Biosensors show high sensitivity and specificity to targets and can be used as simple one-step measurement tools or as multimeasurement devices. Moreover, biosensors can be designed to be operated on-site or at point of care, eliminating the need of expensive lab-based testing. The miniaturization ability of biosensors and their compatibility with data processing technologies, allow them to be integrated into small portable devices. This versatility in biosensors has prompted worldwide research and commercial exploitation

FIGURE 1.1 Schematic representation of a biosensor. 6 MICROBIOLOGICAL DETECTORS FOR FOOD SAFETY APPLICATIONS

120 Forecast

100 PCR

80 Culture methods 60 Biosensors

40 Publications in SCI journals 20 ELISA Gel electrophoresis 0 (b) 1985 1990 1995 2000 2005 2010

FIGURE 1.2 Recent trends in pathogen detection [adapted from Lazcka et al. [22]]. of the technology. Recent trends (Fig. 1.2) indicate that biosensors are the fastest-growing technology for rapid detection of pathogens [22].

1.3 BIOSENSORS FOR MICROBIAL PATHOGEN DETECTION

In this section, we describe different types of biosensors for pathogen detection based on their transduction mechanism such as mechanical, optical, electrochemical, and magnetic approaches.

1.3.1 Mechanical Biosensors 1.3.1.1 Quartz Crystal Microbalance (QCM) Biosensors. Quartz crystal resonators form the basis of Quartz Crystal Microbalance (QCM) sensors. The term “QCM” is used collectively for bulk acoustic wave (BAW), quartz crystal resonance sensors (QCRS), and thickness shear mode (TSM) acoustic sensors [23]. QCM sensors are comprised of a thin quartz disc with electrodes plated on it. When an oscillating electric ﬁeld is applied across the disc, an acoustic wave with a certain resonant frequency is induced. The disc can be coated with a sensing layer of biomolecules based on the analyte to be detected. The interaction of the analyte with the biomolecules on the disc surface causes a change in mass and a concurrent change in resonant frequency that can be directly correlated to the biomolecular interactions [24]. The relation between mass and the resonant frequency is given by the Sauerbrey equation:

−2.3 × 106F 2m F = 0 A (1) where, F is the change in frequency (Hz), F0 is the resonant frequency of the crystal (MHz), m is the deposited mass (grams) and A is the coated area (cm2). The quartz crys- tals are inexpensive, easily available, and robust, thus making them suitable for chemical BIOSENSORS FOR MICROBIAL PATHOGEN DETECTION 7 sensors and biosensors. In addition, QCM-based sensors provide great flexibility, wide dynamic range of frequency measurements, and label-free detection [24]. A wide range of nonlabeled QCM biosensors have been reported in the literature for the detection of pathogenic bacteria and viruses. QCM sensors based on lectin recognition systems for bacterial identification have been studied by Shen et al. [25], Safina et al. [26]. Shen et al. have used a combination of mannose self-assembled monolayer (SAM) and lectin concanavalin A for the detection of E. coli W1485 in a linear range of 7.5 × 102 to 7.5 × 107 cells/ml. Safina et al. utilized lectin reporters to develop a flow injection QCM biosensor for detection of Campylobacter jejuni and Helicobacter pylori. The authors were able to detect 103 to 105 cells/ml in 30 min. A SAM based QCM immunosensor was developed for the detection of E. coli O157:H7 by Su and Li [27]. The immunosensor was able to detect the target bacteria in the range of 103 to 105 CFU/ml in 30–50 min. Detection of B. subtilis spores as a surrogate to B. anthracis was achieved by Lee et al. utilizing a QCM immunosensor to a detection limit of 450 spores/ml [28]. Furthermore, virus (dengue virus and hepatitis B virus) detection with QCM immuno- and nucleic acid- based sensors has been reported by Wu et al. [29] and Yao et al. [30]. QCM biosensors for the detection of DNA sequences have also been developed using nanoparticle labels as amplifiers. Mao et al. [31] reported the use of streptavidin conjugated Fe3O4 nanoparticles (NPs) for the detection of E. coli O157:H7 eaeA gene. The NPs acted as ‘mass enhancers’ and amplified the change in frequency. The biosensor could attain a sensitivity of 10−12 M synthetic oligonucleotides and 2.67 × 102 CFU/ml E. coli O157:H7 cells [31]. Similarly, Au NPs were employed by Wang et al. for real-time bacterial DNA detection in a circulating flow QCM biosensor. The authors reported a sensitivity of 2.0 × 103 CFU/ml for E. coli O157:H7 eaeA gene [32]. A QCM-based biosensor was used to detect Salmonella sp. in milk samples with detection limits around 106 CFU/ml [33]. Tombelli et al. [34] developed a DNA piezoelectric biosensor for the detection of bacterial toxicity based on the detection of PCR amplified aer gene of Aeromonas hydrophila. The biosensor was applied to vegetables, environmental water, and human specimens. The biosensor was able to successfully distinguish between samples containing the pathogen and those not contaminated. Zhao et al. [35] developed a QCM biosensor using 50 nm gold NPs as the amplification probe for DNA detection in the order of 10 fM of target, which was higher than what has been reported using the same method. The high sensitivity was explained by the weight of the larger particles, and the larger area occupied by the larger particles that needed less target DNA for their binding. Another QCM biosensor applied to the detection of E. coli in water in combination with PCR amplification (of the lac gene) was able to detect a 10 fg of genomic E. coli DNA (few viable E. coli cells in 100 ml of water) [36]. When used for detection of Hepatitis B virus, [37] observed that the QCM could detect frequency shifts of DNA hybridization as a linear relationship, in the range 0.02–0.14 μg/ml with a detection limit of 0.1 μg/ml, similar to the QCM biosensor developed by He and Liu [38] for Pseudomonas aeruginosa.

1.3.1.2 Surface Acoustic Wave Biosensors. Surface Acoustic Wave (SAW) sensors are the second class of acoustic wave sensors that have found applications in biosensor devices. SAW sensors consist of two metal interdigital transducers (IDT) etched from a thin metal film deposited on a piezoelectric substrate. The sensing mechanism is based on the changes in SAW velocity or attenuation when mass is sorbed on the sensor surface. Since the acoustic energy is strongly confined to the surface, SAW devices are very 8 MICROBIOLOGICAL DETECTORS FOR FOOD SAFETY APPLICATIONS sensitive to surface changes such as mass loading, viscosity, and conductivity changes [39]. It has been suggested that SAW based biosensors have good sensitivities because of their higher mass sensitivities [39]. SAW biosensors have been successfully applied for the detection of bacteria and viruses. E. coli detection using SAW biosensors have been reported in the literature by multiple authors [40–43]. The biosensors have used antibodies as the biological sensing element with sensitivities ranging from 106 cells/ml to 0.4 cells/μl. Branch and Brozik ◦ have developed a 36 YX-cut LiTaO3 based love-wave device for the detection of the B. anthracis, as simulated by B. thuringiensis spores in aqueous conditions [44]. The authors have investigated two waveguide materials polyimide and polystyrene for creating the love-wave sensors. Detection of B. thuringiensis spores at concentrations below the lethal dose of anthrax spores was possible using both waveguide materials. The sensor had a detection limit of a few hundred cells per ml and a response time of <100 s. Jin et al. [45] developed a SAW biosensor for detecting the gene of Staphylococcal Enterotoxin B utilizing ST-cut quartz and SiO2 guiding layer. The biosensor had a sensitivity of 10 ng/ml and a linear range of 35–200 ng/ml [45]. Recently, SAW biosensors were used for detecting viral bioagents by Bisoffi et al. [46]. A lithium-tantalate based SAW transducer with SiO2 waveguide sensor platform was used for Coxsackie virus B4 and Sin Nomber virus detection. SAW resonators are suitable for use in simple electronic setups because of their low insertion losses and sharp resonance frequencies. As a result, insertion of such devices into oscillator circuits is beneficial as such circuits are commonly used in point-of-care diagnostics. Furthermore, the SAW-based biosensors can be prepared from cheap components thus making them suitable for integration into inexpensive sensor arrays [47].

1.3.1.3 Microcantilever-Based Biosensors. Microcantilever-based biosensors are derived from microfabricated cantilevers used in atomic force microscopy (AFM). Detection is based on the bending induced in the cantilever when a biomolecular interaction takes place on one of its surfaces which is translated into nanomechanical motion and is commonly coupled to an optical or piezoelectric readout system [48]. The cantilevers can be operated in the static deﬂection mode where analyte binding causes cantilever bending or in the dynamic resonant mode where analyte binding causes change in resonant frequency [49]. Microcantilever sensors are promising for biosensor applications since they can perform local, high resolution, and label-free molecular recognition measurements [48]. Davila et al. have demonstrated microcantilever-based biosensors in the detection of B. anthracis Sterne spores in air and water [50]. The detection scheme involved measurement of the decrease in resonant frequency driven by thermally induced oscillations as a result of the mass of spores measured by a laser Doppler vibrometer. The authors reported a minimum detection of 2 spores (740 fg) and 50 spores (139 pg) in air and water, respectively, using 20 μm long, 9 μm wide, and 200 nm thick cantilevers. Campbell and coworkers have utilized piezoelectric-excited millimeter-sized cantilever sensors for the detection of B. anthracis Sterne spores and E. coli O157:H7 cells [51]. The sensors consisted of a piezoelectric and glass layer and were able to detect B. anthracis spores at 300 spores/ml and E. coli O157:H7 cells in ground beef at 50 to 100 cells/ml. Extremely sensitive microcantilever-based biosensors capable of detecting a single pathogen have also been reported in literature [51]. Illic et al. was able to detect a single E. coli O157:H7 cell using low stress silicon nitride cantilever beams in air [52]. The mass of a single E. coli BIOSENSORS FOR MICROBIAL PATHOGEN DETECTION 9

O157:H7 cell was found by the authors to be 665 fg. Similarly, Johnson et al. reported the use of microscale silicon cantilever resonators for vaccinia virus detection in air [53]. The authors measured the mass of a single vaccinia virus particle to be 12.4 ± 1.3fg and 7.9 ± 4.6 fg using two different-sized cantilever beams [51]. The ability to detect small amounts of bacterial organisms was demonstrated using micro-electromechanical systems (MEMS) for the qualitative detection of speciﬁc Salmonella enterica strains with a functionalized silicon nitride microcantilever. Detection was achieved due to a change in the surface stress on the cantilever surface in situ, upon binding of a small number of bacteria with less than 25 adsorbed bacteria required for detection [54].

1.3.2 Optical Biosensors 1.3.2.1 Surface Plasmon Resonance Biosensors. Surface Plasmon Resonance (SPR) is an optical technique for monitoring biomolecular interactions that occur in the close vicinity of a transducer surface. SPR-based biosensing can be subdivided into three categories depending on the mode of SPR detection: angular, spectral, and local SPR biosensing. Angular SPR biosensing is the most common form and involves attenuated total reflection approach using Kretschmann geometry [55]. Spectral SPR biosensing is conducted at a fixed incident angle and utilizes the wavelength dependence of the dielectric constant of the metal film to interrogate the surface plasmon coupling conditions. Local SPR (LSPR) biosensing or nanoparticle-based SPR involves coupling of surface-immobilized metallic NPs or nanostructures into a plasmon mode which results in a decrease in the transmitted power at a specific resonant wavelength dependent on environmental dielectric conditions [55]. SPR biosensors provide several advantages over conventional transduction techniques, such as capability of label-free detection, ability to produce continuous real-time responses, regeneration of the active sensor surface, feasibility for miniaturization, multiplexing ability, and sensitive detection of small molecules [56]. SPR-based immunosensors have been developed for the detection of E. coli O157:H7 by Irudayaraj and coworkers [57, 58]. The authors have demonstrated a sensitivity of 103 CFU/ml for the pathogen using a SAM-based SPR biosensor and the commercially available Spreeta SPR biosensor [57, 58]. Detection of Salmonella Typhimurium in chicken carcasses was achieved using an antibody-based SPR biosensor by Lan et al. [59]. The SPR biosensor had a lowest detection limit of 106 CFU/ml, and highly sensitive detection of Salmonella Enteritidis was attained by Waswa et al. using the commercial BiacoreTM SPR biosensor [60]. The limit of detection (LOD) of the biosensor as reported by the authors was 23 CFU/ml for Salmonella. Chen et al. have demonstrated an immunomagnetic separation based SPR detection method for the foodborne pathogen Staphylococcus aureus [61]. The detection system involved an initial immunomagnetic bead-based separation step of 30 min and had a sensitivity of 106 CFU/ml in a total assay time of 2 h. A SAM-based SPR immunosensor was developed by Jyoung et al. for the detection of V. cholerae O1 with a detection range of 105 to 109 cells/ml [62]. Detection of viruses using SPR-based biosensors have been reported by Chung et al. [63], Vaisocherova et al. [64]. Vaisocherova and coworkers developed an SPR biosensor for detecting antibodies against the Epstein-Barr virus. The antibody detection was performed using an immunoreaction between the antibody and a synthetic peptide for the virus. The sensor had a sensitivity of 0.2 ng/ml (∼1 pM). Furthermore, multiplex detection of four foodborne bacterial pathogens (E. coli O157:H7, Salmonella Typhimurium, 10 MICROBIOLOGICAL DETECTORS FOR FOOD SAFETY APPLICATIONS

Listeria monocytogenes, and Campylobacter jejuni) was demonstrated by Taylor et al. using an eight-channel SPR biosensor [65]. The LOD for each of the four species of bacteria was in the range of 3.4 × 103 and 1.2 × 105 CFU/ml. Additional review chapters focus on different SPR based detection techniques and their applications [56, 66, 67]. Listeria and Salmonella enterica were detected at 106 CFU/ml by an SPR biosensor [68]. Additional SPR biosensors for different bacterial targets, such as P. aeruginosa, B. cereus, and E. coli O157:H7, were later developed by various researchers [27, 69–71], and showed similar detection limits.

1.3.2.2 Fluorescence-Based Biosensors. Fluorescence is the radiative de-excitation of a molecule following the absorption of a photon. Generally, the emitted photon is of lower energy than the absorbed photon, and the fluorescence emission peak of a species is at longer wavelengths than the absorption peak, the wavelength separation being referred to as Stoke’s shift. Fluorescence based detection systems have gained popularity in biosensors due to their high sensitivity and are mostly based on the detection of the fluorescent signal generated by fluorophores that are used to label the biomolecules [72]. Fluorescence detection techniques can be performed in high throughput mode in combination with platforms such as microarrays. Microarrays offer the advantage of using a two-dimensional layout of recognition elements for simultaneous detection and quantifi- cation. Taitt et al. have demonstrated a fluorescence-based microarray immunosensor for the simultaneous detection of nine targets comprising B. anthracis Sterne, B. globigii, Francisella tularensis, Y. pestis, Salmonella Typhimurim, Staphylococcal enterotoxin B, ricin, cholera toxin, and MS2 coliphage [73]. More recently, Li et al. have developed a DNA microarray based on fluorescent nanobarcodes, for the simultaneous detection of DNA of four targets (B. anthracis, Francisella tularensis, Ebola virus, and SARS coronavirus) [74]. The detection procedure involved confocal microscopy, dot blotting and flow cytometry, and resulted in sensitivity in the attomolar range. Fluorescence resonance energy transfer (FRET-based) detection involves nonradiative energy transfer between a donor fluorophore and an acceptor fluorophore when they are in close proximity [75]. A fiber-optic portable biosensor utilizing the principle of FRET was developed by Ko and Grant for rapid detection of Salmonella ser. Typhimurium in ground pork samples [76]. The biosensor had a sensitivity of 105 CFU/ml in a response time of 5 min. Kim et al. reported a molecular beacon (MB) DNA microarray system for fast detection of E. coli O157:H7 based on FRET [77]. In this system, unlike conventional fluorophore-quencher beacon design, two fluorescence molecules allowed active visualization of both hybridized and unhybridized states of the beacon. The target gene detection limit for the system was 1 ng/μl. Fluorescence-based tapered fiber-optic biosensors have also been employed in pathogen detection. A fluorescence-based fiber-optic biosensor for detecting E. coli O157:H7 in ground beef samples was developed by Geng et al. [78]. The authors reported sensitivity of 103 CFU/ml in pure cultures and of 1 CFU/ml in artificially contaminated ground beef samples after 4 h enrichment using a sandwich immunoassay. A fiber-optic biosensor was also developed by Geng et al. for detecting L. monocytogenes in hot-dog and bologna naturally contaminated or artificially inoculated with 10 to 103 CFU/g, after enrichment in buffered Listeria enrichment broth [79]. A cyanine5-labeled antibody was used to generate a specific fluorescent signal. The sensitivity threshold was about 4.3 × 103 CFU/ml. Nanduri et al. developed an automated fiber-optic based immunosensor called RAPTORTM for the detection of Listeria monocytogenes in BIOSENSORS FOR MICROBIAL PATHOGEN DETECTION 11 food samples. The LOD of the system was 5 × 105 CFU/ml in food samples and 1 × 103 CFU/ml in phosphate buffered saline (PBS) solution [80]. In another study, a microcapillary flow injection liposome immunoanalysis system (mFILIA) was developed for the detection of heat-killed E. coli O157:H7. Liposomes tagged with anti-E. coli O157:H7 and an encapsulating fluorescent dye were used to generate fluorescence signals measured by a fluorometer. The mFILIA system successfully detected as few as 360 cells/ml with a total assay time of 45 min [81]. An automated optical biosensor system based on fluorescence excitation and detection in the evanescent field of a quartz fiber was used to detect 16-mer oligonucleotides in DNA hybridization assays. The detection limit for the hybridization with a complementary fluorescein-labeled oligonucleotide was 2 × 10−13 M [82]. Another optical fiber evanescent wave DNA biosensor used an MB-DNA probe that became fluorescent upon hybridization with target DNA [83]. The detection limit of the evanescent wave biosensor with synthesized complementary DNA was 1.1 nM. Liu et al. later developed MB-DNA biosensors with micrometer to submicrometer sizes for DNA/RNA analysis. The MB-DNA biosensor was highly selective with single base-pair mismatch identification capability, and could detect 0.3 nM and 10 nM of rat gamma-actin mRNA with a 105-μm biosensor and a submicrometer (0.1 μm) biosensor, respectively [84]. Optical biosensors targeting RNA as the analyte offer an added advantage over traditional DNA-based detection methods, that is, viable cell detection. Baeumner et al. [85] detected as few as 40 E. coli cells/ml in samples using a simple optical dipstick-type biosensor coupled to Nucleic Acid Sequence Based Amplification (NASBA), emphasizing the fact that only viable cells were detected, and no false positive signals were obtained from dead cells present in a sample. The detection of viable cells is important with respect to safety, and also food and environmental sample sterilization assessments. Similarly, a biosensor for the protozoan parasite Cryptosporidium parvum was developed [86]. Hartley and Baeumner [87] developed a simple membrane-strip based biosensor for the detection of viable B. anthracis spores. The study combined the optical detection process with a spore germination procedure, as well as a nucleic acid amplification reaction to identify as little as one viable B. anthracis sporein12h.

1.3.3 Electrochemical Biosensors Electrochemical biosensors are based on the detection of electrochemical signals generated by consumption or production of electrons from biological interactions occurring at the sensor surface. Advantages such as low cost, high sensitivity, miniaturization ability, low power requirements, and simple instrumentation make electrochemical biosensors well-suited for clinical and environmental analysis. Electrochemical biosensors are generally classiﬁed as amperometric, potentiometric, conductometric, and impedimetric.

1.3.3.1 Amperometric Biosensors. Amperometric biosensors are based on the measurement of current changes resulting from oxidation or reduction of an electroactive species in a biochemical reaction. The current is typically measured at a ﬁxed potential (amperometry) or during controlled variations of the potential (voltammetry). Theegala et al. reported an oxygen-electrode based amperometric biosensor for the qualitative detection of E. coli O157:H7 in water [88]. The biosensor detected changes in oxygen concentration due to decrease in enzymatic activity upon binding of bacterial 12 MICROBIOLOGICAL DETECTORS FOR FOOD SAFETY APPLICATIONS

cells. The biosensor could detect as low as 50 cells/ml in 20 min. A renewable amperometric immunosensor for the detection of S . typhi was reported by Singh et al. [89]. The detection technique involved a sandwich ELISA system with an LOD of 105 cells/ml in 90 min. Amperometric detection of antibodies against B. anthracis protective antigen was also achieved by Aguilar et al. [90]. The antibodies were captured and detected using microcavities with an LOD of 10 fg in a 200 nl sample. A disposable amperometric immunosensor based on a screen-printed electrode (SPE) coated with agarose or nano-Au membrane and horseradish peroxidase-labeled antibody for speciﬁc detection of the foodborne pathogen Vibrio parahaemolyticus was developed by Zhao et al. [91]. The immunosensor showed a sensitivity of 7 × 104 CFU/ml for the pathogen with good consistency (97.5%) as compared to the ELISA results. Lermo et al. described a genomagnetic assay for the electrochemical detection of Salmonella spp. based on in situ DNA ampliﬁcation and magnetic primers [92]. Detec- tion was achieved by double hybridization of the target on magnetic beads which were then separated by a magneto-electrode based on graphite-epoxy composite followed by electrochemical detection using an enzyme marker anti-digoxigenin horseradish peroxidase. The authors achieved a sensitivity of 2.8 fmol with PCR amplicons. Multi-analyte detection using electrochemical genosensors have also been studied by Elsholz et al. [93], Farabullini et al. [94]. Farabullini et al. achieved nanomolar detection limits for the toxin-producing bacteria such as Salmonella sp., E. coli O157:H7, L. monocytogenes, and S. aureus, using differential pulse voltammetry to detect α-napthol signal in less than 1h. Additional biosensors (targeting DNA) that have been developed include MEMS-based amperometric [95] and high throughput PCR biosensors [96], microcantilever-based cyclic voltammetry biosensor [97], pulsed amperometry- [98], and capacitance-based biosensors [99, 100].

1.3.3.2 Potentiometric Biosensors. Potentiometric devices are based on the measurement of accumulation of charge potential at the working electrode of an electrochemical cell in comparison to the reference electrode with zero or negligible current ﬂow between the electrodes. The measured potential is related to the concentration of the analyte through the Nernst equation:

E = E0 ± (RT/nF) ln Q

where, E is the cell potential at zero current, E0 is the standard potential, R is the universal gas constant, T is the absolute temperature, F is the Faraday’s constant, n is the total number of charges of ion, Q is the ratio of ion concentration at the anode to that at the cathode [101, 102]. Among electrochemical transducing methods, potentiometric methods are the least exploited in pathogen detection due to their high detection limits and poor selectiv- ity, the main advantage of these devices being wide detectable concentration range and continuous measurement capability [103]. Another approach involves ion selective ﬁled effect transistors (ISFETs) that employ semiconductor ﬁeld-effect to detect biorecogni- tion events. However, the application of these devices in biosensors has been limited by production problems related to immobilization, fabrication and packaging, poor detection limits, and device stability [22]. An advancement that has evolved from the ISFET is the light addressable potentiometric sensor (LAPS) which combines potentiometry with BIOSENSORS FOR MICROBIAL PATHOGEN DETECTION 13 optical detection [22, 104]. Ercole et al. reported an antibody-based LAPS biosensor for the determination of E. coli in food [105]. The biosensor detected variations in pH due to ammonia production by urease-E. coli antibody conjugates in commercial lettuce, sliced carrot, and rucola samples. The sensor was able to reach a sensitivity of 10 cells per ml in an assay time of 1.5 h.

1.3.3.3 Conductometric Biosensors. Conductometric biosensors utilize the electrical conductivity of a sample to determine the components and their concentration [106]. Muhammad-Tahir and Alocilja [107–110] have developed a conductometric biosensor for the detection of pathogenic bacteria and viruses. The biosensor was fabricated using conducting polyaniline as an electronic label in a sandwich immunoassay scheme, and the authors demonstrated that polyaniline improved the sensitivity of the biosensor by forming a conductive molecular bridge between silver electrodes. The authors reported a sensitivity of 101.8.3 × 101 CFU/ml for Salmonella,7.9 × 101 CFU/ml for E. coli O157:H7, 7.5 × 101 CFU/ml for E. coli, and 103 CCID/ml for BVDV virus in a detection time of 10 min. A conductometric immunosensor based on magnetic NPs has been recently developed by Hnaiein et al.,for the detection of E. coli [111]. The immunosensor was composed of streptavidin-modiﬁed magnetic nanoparticle layer immobilized on a conductometric transducer consisting of interdigitated gold electrodes. Conductivity measurements allowed detection of 0.5 CFU/ml of E. coli without the need for ampliﬁ- cation.

1.3.3.4 Impedimetric Biosensors. Impedance spectroscopy involves applying small amplitude, perturbing sinusoidal voltage signals to an electrochemical cell and measuring the resulting current response. The complex impedance, sum of real and imaginary impedance components, can be calculated as a function of the excitation frequency of the applied potential by varying it over a range of frequencies [112]. Impedimetric detection techniques provide the advantages of high sensitivity, linearized current-potential characteristics, measurement over wide time or frequency ranges, and label-free sensing [22, 106] Radke and Alocilja had developed a microimpedance biosensor for the detection of E. coli O157:H7 [113]. The sensor detected changes in impedance caused by the presence of bacteria immobilized on interdigitated gold electrode arrays fabricated from silicon. The biosensor was able to discriminate between different cellular concentrations of the bacteria (105 to 107 CFU/ml) in 5 min. Nandakumar et al. have demonstrated the detection of Salmonella Typhimurium using electrochemical impedance spectroscopy based on Bayesian decision theory [114]. The technique detected the pathogen in 6 min at a lowest concentration of 500 CFU/ml. An impedance biosensor based on an interdigitated array microelectrode coupled with magnetic nanoparticle-antibody conjugates was developed for rapid and specific detection of E. coli O157:H7 in ground beef samples, by Varshney and Li [115]. Magnitude of impedance and phase angle was measured in a frequency range of 10 Hz to 1 MHz in the presence of 0.1 M mannitol solution. The lowest detection limit of the biosensor for E. coli O157:H7 was 7.4 × 104 CFU/ml in pure cultures and 8.0 × 105 CFU/ml in ground beef samples, the total detection time being 35 min. An impedance biosensor chip for detection of E. coli O157:H7 was developed based on the surface immobilization of affinity-purified antibodies onto indium tin oxide (ITO) electrode chips, with a detection limit of 6 × 103 cells/ml [116]. Shah et al. [117] developed an amperometric immunosensor with a graphite-coated nylon membrane serving as 14 MICROBIOLOGICAL DETECTORS FOR FOOD SAFETY APPLICATIONS a support for antibody immobilization and as a working electrode. This approach was used for detection of E. coli, with a low detection limit of 40 CFU/ml.

1.3.4 Magnetic Biosensors Devices based on the detection of magnetic labels are emerging as a promising new approach in the field of biosensing [118]. Magnetic labels have gained popularity in biosensing because they are physically and chemically stable, are relatively inexpensive, and can easily be made biocompatible. Several approaches have been developed in the past few years for both direct and indirect detection of magnetic labels. Direct detection includes approaches for measuring magnetic parameters, such as magnetic permeability, magnetic remanence, magnetoresistance, and Hall Effect. The indirect detection methods are based on microcantilever-based force amplified sensors and magnetic relaxation switches [119]. However, the applications of these magnetic devices for detection of actual targets such as pathogenic microorganisms remain limited and require further research. Edelstein et al. had developed a multi-analyte BARC (Bead Array Counter) biosensor using giant magnetoresistive (GMR) sensors to detect and identify biological warfare agents [120]. The prototype designed by the authors consisted of a microfabricated chip with GMR sensor arrays, an electronic chip-carrier board, a fluidics cell and an electro- magnet. DNA probes were patterned onto the GMR sensor chips and hybridized with complementary PCR products. Micron-sized magnetic beads were then bound to the DNA sample by streptavidin biotin interactions and the unbound beads were removed by applying a magnetic field. The bound magnetic beads were then detected by the GMR sensors. The authors were able to demonstrate the detection of B. anthracis lethal factor and C. botulinum neurotoxin A using the BARC biosensor. A mass-sensitive magnetoelastic immunosensor for the detection of E. coli O157:H7 was reported by Ruan et al. [121]. The detection was based on the immobilization of alkaline phosphatase-labeled antibodies on the surface of a micrometer-scale magnetoelastic cantilever and amplification of the mass change associated with antigen-antibody binding reaction by biocatalytic precipitation of 5-bromo-4-chloro-3-indolyl phosphate. The minimum detectable level of the immunosensor was 6 × 102 cells/ml. A high efficiency Hall-Effect microbiosensor platform has recently been developed for the detection of magnetically labeled biomolecules by Sandhu et al. [122]. In this system, the integration of Hall-Effect structures with microcurrent lines allowed manipulation of the magnetic beads position via field gradients. The authors studied the hybridization of fully complementary DNA strands of 20–25 bases using Dynabeads as magnetic labels with this platform. Although, the sensitivity of the sensor was not reported, the authors were able to demonstrate a quantitative relationship between the number of magnetic labels and the output signal.

1.4 INTEGRATED EXTRACTION/DETECTION MAGNETIC NANOPARTICLE-BASED BIOSENSOR SYSTEM

In this section, we illustrate an integrated extraction-detection system. Speciﬁcally we present how electrically active polyaniline-coated magnetic (EAPM) NPs are used in a direct charge transfer biosensor for the concentration and detection of B. anthracis INTEGRATED EXTRACTION/DETECTION MAGNETIC NANOPARTICLE-BASED BIOSENSOR 15 spores from complex food matrices such as romaine lettuce, lean ground beef, and ultra-pasteurized whole milk. For lettuce and beef, 25 g samples were weighed, mixed with 225 ml of 0.1% (w/v) peptone water in a Whirl-Pak plastic bag, and stomached in a stomacher (Microbiology International, MD) for one minute. The milk samples were used as purchased. Nine ml of the liquid samples were thoroughly mixed with 1 ml of appropriate concentrations of B. anthracis spores stock solution in a vortex mixer (Fisher Scientiﬁc, IA). Finally, a series of 10 ml samples inoculated with B. anthracis spores at concentrations ranging from 101 to 107 spores/ml were obtained. For details of this illustration, please refer to Pal and Alocilja [123]. The biosensor design is shown in Figure 1.3a [123]. A three-component biosensor system is made up of a sample application pad, capture pad, and absorption pad. Silver electrodes are fabricated along both sides of the capture pad leaving an electrode gap of 0.5 mm. For data acquisition, the biosensor unit is connected to a handheld multimeter linked to a computer. There are three important materials in the detection scheme: detector antibodies conjugated to EAPM-NPs (Abd-EAPM), capture antibodies immobilized on the nitro- cellulose membrane (Abc), and target B. anthracis spores. Figure 1.3b is a schematic representation of the immunomagnetic separation and biosensor detection procedure. The biosensor detection involves a sandwich immunoassay to form the biological structure Abc-spore-Abd-EAPM. To start, EAPM-NPs are conjugated with mouse monoclonal anti-B. anthracis IgG molecules through direct physical adsorption (Step 1), forming Abd-EAPM. Then Abd-EAPM are used to immunomagnetically concentrate the spores

FIGURE 1.3 (a) Biosensor architecture and dimensions; (b) Schematic representation of the biosensor detection system [123]. 16 MICROBIOLOGICAL DETECTORS FOR FOOD SAFETY APPLICATIONS from the complex food matrices (Step 2). The concentrated targets (spore-Abd-EAPM) are then washed to remove unbound materials (Step 3) and applied to the sample application pad of the biosensor (Step 4). The spore-Abd-EAPM complex flows to the capture pad by capillary action, where the antigen is anchored by the capture antibodies (Abc) and the sandwich structure Abc-spore-Abd-EAPM is formed (Step 5). The conductive EAPM-NPs bound to the spores in the sandwich act as a voltage-controlled “ON” switch resulting in a decreased resistance across the silver electrodes that is recorded electrically (step 6) [123]. Figure 1.4a shows the magnetization versus magnetic field i.e. the M-H loop measurements of both unmodified Fe2O3 and synthesized EAPM NPs, using a DC SQUID magnetometer at 300 K. Both systems start to saturate at 5 kOe. The saturation magnetization (M S)fortheFe2O3 NPs is 64.4 emu/g, whereas the saturation magnetization of the EAPM-NPs is 44.1 emu/g. The decrease in M S value for the EAPM-NPs is expected due to surface interactions between the polymer (polyaniline) which is diamagnetic in nature, and iron oxide NPs [124]. The remanent magnetization M r for the Fe2O3 and the EAPM-NPs are 14.2 and 10.4 emu/g, respectively, and the coercive force H c for both NPs is 200 Oe. The low values of M r and H c suggest that the NPs are still in the ferro- magnetic phase but approaching superparamagnetic behavior [125]. Figure 1.4b (inset) shows electrical conductivity measurements of both synthesized EAPM and unmodified Fe2O3 NPs compressed into pellets of 2000 microns thickness at room temperature. The 5 Fe2O3 NPs have an electrical conductivity as low as 3.4/10 S/cm whereas the conductivity of the EAPM-NPs is 5 orders of magnitude higher: 3.3 S/cm. This increase in electrical conductivity is expected and confirms the presence of electrically active polyaniline in the NPs.

1.0E+02 (b) 70 (a)

1.0E+00 50 1.0E-02

1.0E-04 30

Conductivity (S/cm) 1.0E-06 EAPM Fe2O3 10

−15 −5 515 −10

−30

Magnetization (emu/g) Fe2O3 EAPM −50

−70 Field (kOe)

FIGURE 1.4 (a) Experimental M-H curves of the synthesized EAPM and unmodiﬁed Fe2O3 nanoparticles at 300 K; and (b, inset) Electrical conductivity measurements for the EAPM and Fe2O3 nanoparticles at room temperature [printed with permission from Pal and Alocilja [123]]. INTEGRATED EXTRACTION/DETECTION MAGNETIC NANOPARTICLE-BASED BIOSENSOR 17

(a) (b)

FIGURE 1.5 TEM images of (a) unmodiﬁed Fe2O3 nanoparticles; and (b) Synthesized EAPM nanoparticles [printed with permission from Pal and Alocilja [123]].

Figure 1.5a and 1.5b reveal Transmission Electron Microscope (TEM) images of the unmodified Fe2O3 NPs and the polyaniline-coated EAPM-NPs. The iron oxide NPs have an average diameter of 20 nm according to manufacturer’s specifications, which is consistent with the TEM image in Figure 1.5a, whereas the polyaniline-coated EAPM-NPs show a diameter ranging from 50 to 100 nm (Figure 1.5b). The immunomagnetic capture of B. anthracis spores using EAPM-NPs from the three different food matrices was confirmed by microbial plating in TSA II blood agar plates and the capture efficacy of the NPs was evaluated. The capture ratio (CR) of the EAPM-NPs was evaluated using the formula CR = Ccaptured/Cactual, where, C actual is the actual concentration of viable spores in the sample, and C captured is the concentration of viable spores extracted from the food samples. The capture effects of the EAPM-NPs on B. anthracis spores from the lettuce and ground beef samples are quite similar. In both samples, the capture effect is highest at a viable spore concentration of 1.52 × 102 CFU/ml with a CR value of 0.97 for lettuce and that of 0.72 for ground beef. Also, the EAPM-NPs could achieve an immunomagnetic capture at viable spore concentration as low as 1.52 × 101 CFU/ml from both lettuce and ground beef samples. However, for whole milk samples, the immunomagnetic capture of the viable B. anthracis spores could only go as low as 1.52 × 102 CFU/ml and the capture effects are observed to be similar for all viable spore concentrations with CR values in the range of 0.06 and 0.11. This could be explained by the high fat content of the whole milk samples which might have interfered in the immunomagnetic capture process. Figure 1.6a, b, and c show the average resistance readings measured with the EAPM nanoparticle-based direct charge transfer biosensor in lettuce, whole milk, and ground beef samples inoculated with B. anthracis spores. The average resistance signals obtained from three replicates were plotted for the control and the food samples, contaminated with spore concentrations ranging from 4.2 × 101 to 4.2 × 107 spores/ml. As observed, the resistance values recorded for the different spore concentrations are much lower than the values for the control solution that has no spores in it. The average resistances for the control solutions for all three food samples are in the range of 288 ± 50 and 353 ± 51 k, whereas the average resistances for the different spore concentrations in the three food samples vary from 75.1 ± 14 to 132.6 ± 19 k. The reduced resistance values support the 18 MICROBIOLOGICAL DETECTORS FOR FOOD SAFETY APPLICATIONS

500 500 (a) (b) a a 400 a 400 a ) ) Ω Ω 300 300

c 200 c b,c 200 c bb b c b,c Resistance (k

Resistance (k b b 100 100

0 0 Control 10^ 7 10^ 6 10^ 5 10^ 4 10^ 3 10^ 2 10^ 1 Control 10^ 7 10^ 6 10^ 5 10^ 4 10^ 3 10^ 2 Spore concentration/ml Spore concentration/ml

500 (c) a 400 a ) Ω

300

200 c b,c c Resistance (k b b b 100

0 Control 10^ 7 10^ 6 10^ 5 10^ 4 10^ 3 10^ 2 10^ 1 Spore concentration/ml FIGURE 1.6 The EAPM nanoparticle-based direct charge transfer biosensor resistance responses in (a) Romaine lettuce, (b) Whole milk, and (c) Ground beef samples contaminated with B. anthracis spores. Average resistances with the same letters are not significantly different (P > 0.05). formation of a sandwich complex on the capture pad, where the conductive EAPM-NPs act as a charge transfer agent causing a drop in the resistance signal across the silver electrodes [126–129]. Single factor analysis of variance (ANOVA) to a significance of 95% (P < 0.05) was used to compare the differences in the resistance values between the control and the different spore concentrations. The lowest spore concentration that produced a resistance signal significantly different (P < 0.05) from the control was considered to be the sensitivity or detection limit of the biosensor. For the lettuce and ground beef samples, the biosensor sensitivity was 4.2 × 102 spores/ml with statistically significant differences from the control (P-value for lettuce at 102 spores/ml was 1.79 E-05; P-value for ground beef at 102 spores/ml was 2.63E-06). For whole milk samples, the biosensor could reach a sensitivity of 4.2 × 103 spores/ml where statistically significant differences could be observed from the control (P-value at 103 spores/ml was 8.47E-08). The reduced biosensor sensitivity in the whole milk samples could be attributed to the high fat content in these samples. As observed in Figure 1.6, although the biosensor resistance readings recorded for the different spore concentrations were different from the control, statistical analysis did not reveal any significant differences between the concentrations. Artifacts in biosensor fabrication, probabilistic antigen-antibody interactions, antibody orientations, and stability of the sandwich complex on the capture pad might be some of the factors behind such biosensor performance. At this stage the biosensor is only considered to be a qualitative device for a yes/no diagnosis of B. anthracis spores. However, the biosensor shows excellent sensitivity and fast detection time in comparison to the very few rapid detection systems for B. anthracis in the food matrices that have been reported in the literature [130, 131]. REFERENCES 19

Specificity evaluation of the biosensor is also presented here. A comparison of the biosensor resistance responses was made in pure cultures of E. coli with cell concentrations ranging from 1.7 × 101 to 1.7 × 105 CFU/ml, in pure cultures of Salmonella Enteritidis with cell concentrations ranging from 1.6 × 101 to 1.6 × 105 CFU/ml, and pure spore suspensions of B. anthracis with spore concentrations ranging from 4.2 × 101 to 4.2 × 105 spores/ml. The biosensor average resistance values for different concentrations of the nontarget bacteria (i.e. E. coli and Salmonella Enteritidis) are similar to the values observed for the control. Single factor ANOVA tests to a significance of 95% (P < 0.05) showed no statistically significant differences between the control and different cell concentrations of E. coli and Salmonella Enteritidis with P-values ranging from 0.278 to 0.887 for E. coli, and from 0.348 to 0.981 for Salmonella Enteritidis. The results indicate that the effects of nonspecific interactions are not significant for the range of cell concentrations tested on the biosensor. In comparison, for pure B. anthracis spore suspensions, the biosensor average resistance responses show significant differences between the control and spore concentrations ranging from 102 to 105 spore/ml (P-value range: 0.009−0.0009) which is expected since the antibodies used in the biosensor are specific for B. anthracis.

1.5 CONCLUDING COMMENTS

In this chapter, we attempted to present biosensors using various transduction mechanisms that have been developed for rapid detection of microbial pathogens of concern to food defense and food safety. These biosensors are designed for rapid, highly sensitive, speciﬁc, and user-friendly operation. While they are not exhaustive, the chapter provides a wide range and scope of the detection mechanisms that are novel and potentially market-ready. The illustrated biosensor on the EAPM-based system is an excellent demonstration on the potential speed, sensitivity, and speciﬁcity that can be achieved by biosensors in general.

REFERENCES

1. Inglesby, T. V., Dennis, D. T., Henderson, D. A., Bartlett, J. G., Ascher, M. S., Eitzen, E., et al. (2000). Plague as a biological weapon-medical and public health management. JAMA 283(17), 2281–2290. 2. Jernigan, J. A., Stephens, D. S., Ashford, D. A., Omenaca, C., Topiel, M. S., Galbraith, M., et al. (2001). Bioterrorism-related inhalational anthrax: the ﬁrst 10 cases reported in the United States. Emerging Infect. Dis. 7(6), 933–944. 3. Spencer, R. C. (2003). Bacillus anthracis. J. Clin. Pathol. 56(3), 182–187. 4. Mock, M., and Fouet, A. (2001). Anthrax. Annu. Rev. Microbiol. 55, 647–671. 5. Okinaka, R. T., Cloud, K., Hampton, O., Hoffmaster, A. R., Hill, K. K., Keim, P., et al. (1999). Sequence and organization of pXO1, the large Bacillus anthracis plasmid harboring the anthrax toxin genes. J. Bacteriol. 181(20), 6509–6515. 6. Collier, R. J., and Young, J. A. T. (2003). Anthrax toxin. Annu. Rev. Cell Dev. Biol. 19, 45–70. 7. Sirisanthana, T., and Brown, A. E. (2002). Anthrax of the gastrointestinal tract. Emerging Infect. Dis. 8(7), 649–651. 20 MICROBIOLOGICAL DETECTORS FOR FOOD SAFETY APPLICATIONS

8. Mock, M., and Mignot, T. (2003). Anthrax toxins and the host: a story of intimacy. Cell. Microbiol. 5(1), 15–23. 9. Mead, P. S., Slutsker, L., Dietz, V., McGaig, L., Bresee, J., Shapiro, C., Griffin, P., and Tauxe, R. (1999). Food-related illnesses and death in the United States. Emerging Infect. Dis. 5, 607–625. 10. CDC. (2001a). Outbreaks Caused by Shiga Toxin-producing Escherichia Coli-Summary of 2000 Surveillance Data. Centers for Disease Control and Prevention. Available at http://www.cdc.gov/foodborneoutbreaks/ecoli/2000 summaryLetter.pdf. 11. Doyle, M. P., Zhao, T., Meng, J., and Zhao, S. (1997). Escherichia coli O157:H7. Food Microbiology Fundamentals and Frontiers. American Society for Microbiology, Washington, DC. 12. FDA. (2006). Foodborne Pathogenic Microorganisms and Natural Toxins Handbook: The “Bad Bug Book”. FDA-CFSAN . Available at http://www.cfsan.fda.gov/∼mow/intro.html 13. WHO. (2002). Terrorist Threats to Food: Guidance for Establishing and Strengthening Pre- vention and Response Systems. World Health Organization Food Safety Dept, Geneva, Switzerland. 14. CDC. (2009). Multistate Outbreak of E. coli O157:H7 Infections Linked to Eating Raw Refrigerated, Prepackaged Cookie Dough. Updated June 25, 2009 . Available at http://www. cdc.gov/ecoli/2009/0619.html 15. FDA. (2005). Bacteriological Analytical Manual. Food and Drug Administration, Rockville, MD. Available at http://www.cfsan.fda.gov/∼ebam/bam-toc.html 16. CDC. (2002b). Preliminary FoodNet Data on the Incidence of Foodborne Illnesses-Selected Sites, United States, 2001. MMWR 51: 325-9 . 17. CDC. (2002a). Notice to Readers: Final 2001 Reports of Notifiable Diseases. MMWR 51: 710 . 18. CDC. (2001b). Salmonellosis. Available at http://www.cdc.gov/ncidod/dbmd/diseaseinfo/ salmonellosis g.htm 19. FDA. (1992). Foodborne Pathogenic Microorganisms and Natural Toxins Handbook: Salmonella spp. Available at http://www.cfsan.fda.gov/∼mow/chap1.html 20. Inglesby, T. V. (2000). Anthrax as a biological weapon: medical and public health management (vol 281, pg 1735, 1999). JAMA 283(15), 1963. 21. Erickson, M. C., and Kornacki, J. L. (2003). Bacillus anthracis: current knowledge in relation to contamination of food. J. Food Prot. 66(4), 691–699. 22. Lazcka, O., Del Campo, F. J., and Munoz, F. X. (2007). Pathogen detection: a perspective of traditional methods and biosensors. Biosens. Bioelectron. 22(7), 1205–1217. 23. Cooper, M. A., and Singleton, V. T. (2007). A survey of the 2001 to 2005 quartz crystal microbalance biosensor literature: applications of acoustic physics to the analysis of biomolecular interactions. J. Mol. Recognit. 20(3), 154–184. 24. O’Sullivan, C. K., and Guilbault, G. G. (1999). Commercial quartz crystal microbalances-theory and applications. Biosens. Bioelectron. 14(8–9), 663–670. 25. Shen, Z. H., Huang, M. C., Xiao, C. D., Zhang, Y., Zeng, X. Q., and Wang, P. G. (2007). Nonlabeled quartz crystal microbalance biosensor for bacterial detection using carbohydrate and lectin recognitions. Anal. Chem. 79(6), 2312–2319. 26. Safina, G., van Lier, M., and Danielsson, B. (2008). Flow-injection assay of the pathogenic bacteria using lectin-based quartz crystal microbalance biosensor. Talanta 77(2), 468–472. 27. Su, X. L., and Li, Y. B. (2004). A self-assembled monolayer-based piezoelectric immunosensor for rapid detection of Escherichia coli O157: H7. Biosens. Bioelectron. 19(6), 563–574. REFERENCES 21

28. Lee, S. H., Stubbs, D. D., Cairney, J., and Hunt, W. D. (2005). Rapid detection of bacterial spores using a quartz crystal microbalance (QCM) immunoassay. IEEE Sens. J. 5(4), 737–743. 29. Wu, T. Z., Su, C. C., Chen, L. K., Yang, H. H., Tai, D. F., and Peng, K. C. (2005). Piezoelectric immunochip for the detection of dengue fever in viremia phase. Biosens. Bioelectron. 21(5), 689–695. 30. Yao, C. Y., Zhu, T. Y., Tang, J., Wu, R., Chen, Q. H., Chen, M., et al. (2008). Hybridization assay of hepatitis B virus by QCM peptide nucleic acid biosensor. Biosens. Bioelectron. 23(6), 879–885. 31. Mao, X. L., Yang, L. J., Su, X. L., and Li, Y. B. (2006). A nanoparticle amplification based quartz crystal microbalance DNA sensor for detection of Escherichia coli O157: H7. Biosens. Bioelectron. 21(7), 1178–1185. 32. Wang, L. J., Wei, Q. S., Wu, C. S., Hu, Z. Y., Ji, J., and Wang, P. (2008). The Escherichia coli O157:H7 DNA detection on a gold nanoparticle-enhanced piezoelectric biosensor. Chin. Sci. Bull. 53(8), 1175–1184. 33. Park, I. S., Kim, W. Y., and Kim, N. (2000). Operational characteristics of an antibody-immobilized QCM system detecting Salmonella spp. Biosens. Bioelectron. 15, 167–172. 34. Tombelli, S., Mascini, M., Sacco, C., and Turner, A. P. F. (2000). A DNA piezoelectric biosensor assay coupled with a polymerase chain reaction for bacterial toxicity determination in environmental samples. Anal. Chim. Acta 418,1–9. 35. Zhao, H. Q., Lin, L., Li, J. R., Tang, J. A., Duan, M. X., and Jiang, L. (2001). DNA biosensor with high sensitivity amplified by gold nanoparticles. J. Nanopart. Res. 3, 321–323. 36. Mo, X. T., Zhou, Y. P., Lei, H., and Deng, L. (2002). Microbalance-DNA probe method for the detection of specific bacteria in water. Enzyme Microb. Technol. 30, 583–589. 37. Zhou, X. D., Liu, L. J., Hu, M., Wang, L. L., and Hu, J. M. (2002). Detection of Hepatitis B virus by piezoelectric biosensor. J. Pharm. Biomed. Anal. 27, 341–345. 38. He, F. J., and Liu, S. Q. (2004). Detection of P. aeruginosa using nano-structured electrode-separated piezoelectric DNA biosensor. Talanta 62, 271–277. 39. Galipeau, D. W., Story, P. R., Vetelino, K. A., and Mileham, R. D. (1997). Surface acoustic wave microsensors and applications. Smart Mater. Struct. 6(6), 658–667. 40. Berkenpas, E., Millard, P., and da Cunha, M. P. (2006). Detection of Escherichia coli O157:H7 with langasite pure shear horizontal surface acoustic wave sensors. Biosens. Bio- electron. 21(12), 2255–2262. 41. Deobagkar, D. D., Limaye, V., Sinha, S., and Yadava, R. D. S. (2005). Acoustic wave immunosensing of Escherichia coli in water. Sens. Actuators, B Chem. 104(1), 85–89. 42. Moll, N., Pascal, E., Dinh, D. H., Pillot, J. P., Bennetau, B., Rebiere, D., et al. (2007). A Love wave immunosensor for whole E-coli bacteria detection using an innovative two-step immobilisation approach. Biosens. Bioelectron. 22(9–10), 2145–2150. 43. Moll, N., Pascal, E., Dinh, D. H., Lachaud, J. L., Vellutini, L., Pillot, J. P., et al. (2008). Multipurpose Love acoustic wave immunosensor for bacteria, virus or proteins detection. Irbm 29(2–3), 155–161. 44. Branch, D. W., and Brozik, S. M. (2004). Low-level detection of a Bacillus anthracis sim- ulant using Love-wave biosensors on 36 degrees YX LiTaO3. Biosens. Bioelectron. 19(8), 849–859. 45. Jin, X., Gao, Z., Pan, H., Zhu, H., Zhou, M., and Chen, H. (2003). The surface acoustic wave biosensor for detecting the gene of Staphylococal Enterotoxin B. Proceedings of the International Symposium on Test and Measurement 1 , 261–264. 22 MICROBIOLOGICAL DETECTORS FOR FOOD SAFETY APPLICATIONS

46. Bisoffi, M., Hjelle, B., Brown, D. C., Branch, D. W., Edwards, T. L., Brozik, S. M., et al. (2008). Detection of viral bioagents using a shear horizontal surface acoustic wave biosensor. Biosens. Bioelectron. 23(9), 1397–1403. 47. Lange, K., Rapp, B. E., and Rapp, M. (2008). Surface acoustic wave biosensors: a review. Anal. Bioanal. Chem. 391(5), 1509–1519. 48. Carrascosa, L. G., Moreno, M., Alvarez, M., and Lechuga, L. M. (2006). Nanomechanical biosensors: a new sensing tool. Trends Analyt. Chem. 25(3), 196–206. 49. Waggoner, P. S., and Craighead, H. G. (2007). Micro- and nanomechanical sensors for environmental, chemical, and biological detection. Lab Chip 7(10), 1238–1255. 50. Davila, A. P., Jang, J., Gupta, A. K., Walter, T., Aronson, A., and Bashir, R. (2007). Microres- onator mass sensors for detection of Bacillus anthracis Sterne spores in air and water. Biosens. Bioelectron. 22(12), 3028–3035. 51. Campbell, G. A., and Mutharasan, R. (2006). Piezoelectric-excited millimeter-sized cantilever (PEMC) sensors detect Bacillus anthracis at 300 spores/mL. Biosens. Bioelectron. 21(9), 1684–1692. 52. Ilic, B., Czaplewski, D., Zalalutdinov, M., Craighead, H. G., Neuzil, P., Campagnolo, C., and Batt, C. (2001). Single cell detection with micromechanical oscillators. J. Vac. Sci. Technol. B 19(6), 2825–2828. 53. Johnson, L., Gupta, A. T. K., Ghafoor, A., Akin, D., and Bashir, R. (2006). Characterization of vaccinia virus particles using microscale silicon cantilever resonators and atomic force microscopy. Sens. Actuators, B Chem. 115(1), 189–197. 54. Weeks, B. L., Camarero, J., Noy, A., Miller, A. E., Stanker, L., and De Yoreo, J. J. (2003). A microcantilever-based pathogen detector. Scanning 25, 297–299. 55. Erickson, D., Mandal, S., Yang, A. H. J., and Cordovez, B. (2008). Nanobiosensors: optofluidic, electrical and mechanical approaches to biomolecular detection at the nanoscale. Microfluid. Nanofluidics 4(1–2), 33–52. 56. Shankaran, D. R., Gobi, K. V. A., and Miura, N. (2007). Recent advancements in surface plasmon resonance immunosensors for detection of small molecules of biomedical, food and environmental interest. Sens. Actuators, B Chem. 121(1), 158–177. 57. Waswa, J., Irudayaraj, J., and DebRoy, C. (2007). Direct detection of E-coli O157:H7 in selected food systems by a surface plasmon resonance biosensor. LWT-Food Sci. Technol. 40(2), 187–192. 58. Subramanian, A., Irudayaraj, J., and Ryan, T. (2006). A mixed self-assembled monolayer- based surface plasmon immunosensor for detection of E-coli O157: H7. Biosens. Bioelectron. 21(7), 998–1006. 59. Lan, Y. B., Wang, S. Z., Yin, Y. G., Hoffmann, W. C., and Zheng, X. Z. (2008). Using a surface plasmon resonance biosensor for rapid detection of Salmonella typhimurium in chicken carcass. J. Bionic Eng. 5(3), 239–246. 60. Waswa, J. W., DebRoy, C., and Irudayaraj, J. (2006). Rapid detection of Salmonella enteritidis and Escherichia coli using surface plasmon resonance biosensor. J. Food Process Eng. 29(4), 373–385. 61. Chen, L. L., Deng, L., Liu, L. L., and Peng, Z. H. (2007). Immunomagnetic separation and MS/SPR end-detection combined procedure for rapid detection of Staphylococcus aureus and protein A. Biosens. Bioelectron. 22(7), 1487–1492. 62. Jyoung, J. Y., Hong, S. H., Lee, W., and Choi, J. W. (2006). Immunosensor for the detection of Vibrio cholerae O1 using surface plasmon resonance. Biosens. Bioelectron. 21(12), 2315–2319. 63. Chung, J. W., Kim, S. D., Bernhardt, R., and Pyun, J. C. (2005). Application of SPR biosensor for medical diagnostics of human hepatitis B virus (hHBV). Sens. Actuators, B Chem. 111, 416–422. REFERENCES 23

64. Vaisocherova, H., Mrkvova, K., Piliarik, M., Jinoch, P., Steinbachova, M., and Homola, J. (2007). Surface plasmon resonance biosensor for direct detection of antibody against Epstein-Barr virus. Biosens. Bioelectron. 22(6), 1020–1026. 65. Taylor, A. D., Ladd, J., Yu, Q., Chen, S., Homola, J., and Jiang, S. (2006). Quantitative and simultaneous detection of four foodborne bacterial pathogens with a multi-channel SPR sensor. Biosens. Bioelectron. 22(5), 752–758. 66. Homola, J. (2008). Surface plasmon resonance sensors for detection of chemical and biological species. Chem. Rev. 108(2), 462–493. 67. Hoa, X. D., Kirk, A. G., and Tabrizian, M. (2007). Towards integrated and sensitive surface plasmon resonance biosensors: a review of recent progress. Biosens. Bioelectron. 23, 151–160. 68. Koubova, V., Brynda, E., Karasova, L., Skvor, J., Homola, J., Dostalek, J., Tobiska, P., and Rosicky, J. (2001). Detection of foodborne pathogens using surface plasmon resonance biosensors. Sens. Actuators, B Chem. 74, 100–105. 69. Vaughan, R. D., Carter, R. M., O’Sullivan, C. K., and Guilbault, G. G. (2003). A quartz crystal microbalance (QCM) sensor for the detection of Bacillus cereus. Anal. Lett. 36, 731–747. 70. Kim, N., Park, I. S., and Kim, D. K. (2004). Characteristics of a label-free piezoelectric immunosensor detecting Pseudomonas aeruginosa. Sens. Actuators, B Chem. 100, 432–438. 71. Su, X. L., and Li, Y. (2005). Surface plasmon resonance and quartz crystal microbalance immunosensors for detection of Escherichia coli O157: H7. Trans. ASAE 48, 405–413. 72. Zhang, D., Carr, D. J., and Alocilja, E. C. (2009). Fluorescent bio-barcode DNA assay for the detection of Salmonella enterica serovar Enteritidis. Biosens. Bioelectron. 24(5), 1377–1381. 73. Taitt, C. R., Anderson, G. P., Lingerfelt, B. M., Feldstein, M. J., and Ligler, F. S. (2002). Nine-analyte detection using an array-based biosensor. Anal. Chem. 74(23), 6114–6120. 74. Li, Y. G., Cu, Y. T. H., and Luo, D. (2005). Multiplexed detection of pathogen DNA with DNA-based fluorescence nanobarcodes. Nat. Biotechnol. 23(7), 885–889. 75. Epstein, J. R., Biran, I., and Walt, D. R. (2002). Fluorescence-based nucleic acid detection and microarrays. Anal. Chim. Acta 469(1), 3–36. 76. Ko, S. H., and Grant, S. A. (2006). A novel FRET-based optical fiber biosensor for rapid detection of Salmonella Typhimurium. Biosens. Bioelectron. 21(7), 1283–1290. 77. Kim, H., Kane, M. D., Kim, S., Dominguez, W., Applegate, B. M., and Savikhin, S. (2007). A molecular beacon DNA microarray system for rapid detection of E-coli O157:H7 that eliminates the risk of a false negative signal. Biosens. Bioelectron. 22(6), 1041–1047. 78. Geng, T., Uknalis, J., Tu, S. I., and Bhunia, A. K. (2006). Fiber-optic biosensor employing Alexa-Fluor conjugated antibody for detection of Escherichia coli O157: H7 from ground beef in four hours. Sensors 6(8), 796–807. 79. Geng, T., Morgan, M. T., and Bhunia, A. K. (2004). Detection of low levels of Listeria monocytogenes cells by using a fiber-optic immunosensor. Appl. Environ. Microbiol. 70, 6138–6146. 80. Nanduri, V., Kim, G., Morgam, M. T., Ess, D., Hahm, B., Kothapalli, A., et al. (2006). Antibody immobilization on waveguides using a flow-through system shows improved Liste- ria monocytogenes detection in an automated fiber optic biosensor: RAPTOR™. Sensors 6, 808–822. 81. Ho, J.-A. A., Hsu, H.-W., and Huang, M.-R. (2004). Liposome-based microcapillary immunosensor for detection of Escherichia coli O157:H7. Anal. Biochem. 330, 342–349. 24 MICROBIOLOGICAL DETECTORS FOR FOOD SAFETY APPLICATIONS

82. Abel, A. P., Weller, M. G., Duveneck, G. L., Ehrat, M., and Widmer, H. M. (1996). Fiber-optic evanescent wave biosensor for the detection of oligonucleotides. Anal. Chem. 68, 2905–2912. 83. Liu, X., and Tan, W. (1999). A fiber-optic evanescent wave DNA biosensor based on novel molecular beacons. Anal. Chem. 71, 5054–5059. 84. Liu, C. H., Liao, K. T., and Huang, H. J. (2000). Amperometric immunosensors based on protein A coupled polyaniline-perfluorosulfonated ionomer composite electrodes. Anal. Chem. 72, 2925–2929. 85. Baeumner, A. J., Cohen, R. N., Miksic, V., and Min, J. (2003). RNA biosensor for the rapid detection of viable Escherichia coli in drinking water. Biosens. Bioelectron. 18, 405–413. 86. Esch, M. B., Locascio, L. E., Tarlov, M. J., and Durst, R. A. (2001). Detection of viable Cryptosporidium parvum using DNA-modified liposomes in a microfluidic chip. Anal. Chem. 73, 2952–2958. 87. Hartley, H. A., and Baeumner, A. J. (2003). Biosensor for the specific detection of a single viable B. anthracis spore. Anal. Bioanal. Chem. 376, 319–327. 88. Theegala, C. S., Small, D. D., and Monroe, W. T. (2008). Oxygen electrode-based single antibody amperometric biosensor for qualitative detection of E-coli and bacteria in water. J. Environ. Sci. Health A Tox. Hazard Subst. Environ. Eng. 43(5), 478–487. 89. Singh, C., Agarwal, G. S., Rai, G. P., Singh, L., and Rao, V. K. (2005). Specific detection of Salmonella typhi using renewable amperometric immunosensor. Electroanalysis 17(22), 2062–2067. 90. Aguilar, Z. P., and Sirisena, M. (2007). Development of automated amperometric detection of antibodies against Bacillus anthracis protective antigen. Anal. Bioanal. Chem. 389(2), 507–515. 91. Zhao, G., Xing, F., and Deng, S. (2007). A disposable amperometric enzyme immunosensor for rapid detection of Vibrio parahaemolyticus in food based on agarose/Nano-Au membrane and screen-printed electrode. Electrochem. Commun. 9(6), 1263–1268. 92. Lermo, A., Campoy, S., Barbe, J., Hernandez, S., Alegret, S., and Pividori, M. (2007). In situ DNA amplification with magnetic primers for the electrochemical detection of food pathogens. Biosens. Bioelectron. 22(9–10), 2010–2017. 93. Elsholz, B., Worl, R., Blohm, L., Albers, J., Feucht, H., Grunwald, T., et al. (2006). Automated detection and quantitation of bacterial RNA by using electrical microarrays. Anal. Chem. 78(14), 4794–4802. 94. Farabullini, F., Lucarelli, F., Palchetti, I., Marrazza, G., and Mascini, M. (2007). Dispos- able electrochemical genosensor for the simultaneous analysis of different bacterial food contaminants. Biosens. Bioelectron. 22(7), 1544–1549. 95. Gau, J.-J., Lan, E. H., Dunn, B., Ho, C.-M., and Woo, J. C. S. (2001). A MEMS based amperometric detector for E. coli bacteria using self-assembled monolayers. Biosens. Bioelectron. 16, 745–755. 96. Nagai, H., Murakami, Y., Yokoyama, K., and Tamiya, E. (2001). High-throughput PCR in silicon based microchamber array. Biosens. Bioelectron. 16, 1015–1019. 97. Zhang, Z. X., and Li, M. Q. (2005). Electrostatic microcantilever array biosensor and its application in DNA detection. Prog. Biochem. Biophys. 32, 314–317. 98. Ramanaviciene, A., and Ramanavicius, A. (2004). Pulsed amperometric detection of DNA with an ssDNA/polypyrrole-modified electrode. Anal. Bioanal. Chem. 379, 287–293. 99. Berney, H., West, J., Haefele, E., Alderman, J., Lane, W., and Collins, J. K. (2000). A DNA diagnostic biosensor: development, characterisation and performance. Sens. Actuators, B Chem. 68, 100–108. 100. Lee, J. S., Choi, Y.-K., Pio, M., Seo, J., and Lee, L. P. (2002). Nanogap capacitors for label free DNA analysis. BioMEMS Bionanotechnol. 729, 185–190. REFERENCES 25

101. Diamond, D. (1998). Principles of Chemical and Biological Sensors. John Wiley & Sons, New York. 102. Eggins, B. R. (2002). Chemical Sensors and Biosensors. John Wiley & Sons, Chichester. 103. Palchetti, I., and Mascini, M. (2008). Electroanalytical biosensors and their potential for food pathogen and toxin detection. Anal. Bioanal. Chem. 391(2), 455–471. 104. Hafeman, D. G., Parce, J. W., and Mcconell, H. M. (1988). Light-addressable potentiometric sensor for biochemical systems. Science 240(4856), 1182–1185. 105. Ercole, C., Del Gallo, M., Mosiello, L., Baccella, S., and Lepidi, A. (2003). Escherichia coli detection in vegetable food by a potentiometric biosensor. Sens. Actuators, B Chem. 91(1–3), 163–168. 106. Rahman, M. A., Kumar, P., Park, D. S., and Shim, Y. B. (2008). Electrochemical sensors based on organic conjugated polymers. Sensors 8(1), 118–141. 107. Muhammad-Tahir, Z., and Alocilja, E. C. (2003a). A conductometric biosensor for biosecurity. Biosens. Bioelectron. 18(5–6), 813–819. 108. Muhammad-Tahir, Z., and Alocilja, E. C. (2003b). Fabrication of a disposable biosensor for Escherichia coli O157:H7 detection. IEEE Sens. J. 3, 345–351. 109. Muhammad-Tahir, Z., Alocilja, E. C., and Grooms, D. L. (2005a). Polyaniline synthesis and its biosensor application. Biosens. Bioelectron. 20, 1690–1695. 110. Muhammad-Tahir, Z., Alocilja, E. C., and Grooms, D. L. (2005b). Rapid detection of Bovine viral diarrhea virus as surrogate of bioterrorism agents. IEEE Sens. J. 5(4), 757–762. 111. Hnaiein, M., Hassen, W. M., Abdelghani, A., Fournier-Wirth, C., Coste, J., Bessueille, F., et al. (2008). A conductometric immunosensor based on functionalized magnetite nanoparticles for E. coli detection. Electrochem. Commun. 10(8), 1152–1154. 112. Katz, E., and Willner, I. (2003). Probing biomolecular interactions at conductive and semi- conductive surfaces by impedance spectroscopy: routes to impedimetric immunosensors, DNA-Sensors, and enzyme biosensors. Electroanalysis 15(11), 913–947. 113. Radke, S. M., and Alocilja, E. C. (2005). A high density microelectrode array biosensor for detection of E. coli O157:H7. Biosens. Bioelectron. 20(8), 1662–1667. 114. Nandakumar, V., La Belle, J. T., Reed, J., Shah, M., Cochran, D., Joshi, L., and Alford, T. L. (2008). A methodology for rapid detection of Salmonella Typhimurium using label-free electrochemical impedance spectroscopy. Biosens. Bioelectron. 24(4), 1039–1042. 115. Varshney, M., and Li, Y. (2007). Interdigitated array microelectrode based impedance biosensor coupled with magnetic nanoparticle-antibody conjugates for detection of Escherichia coli O157:H7 in food samples. Biosens. Bioelectron. 22(11), 2408–2414. 116. Ruan, C. M., Yang, L. J., and Li, Y. B. (2002). Immunobiosensor chips for detection of Escherichia coli O157: H7 using electrochemical impedance spectroscopy. Anal. Chem. 74, 4814–4820. 117. Shah, J., Chemburu, S., Wilkins, E., and Abdel-Hamid, I. (2003). Rapid amperometric immunoassay for Escherichia coli based on graphite coated nylon membranes. Electroanalysis 15, 1809–1814. 118. Wang, S. X., and Li, G. (2008). Advances in giant magnetoresistance biosensors with magnetic nanoparticle tags: review and outlook. IEEE Trans. Magn. 44(7), 1687–1702. 119. Tamanaha, C. R., Mulvaney, S. P., Rife, J. C., and Whitman, L. J. (2008). Magnetic labeling, detection, and system integration. Biosens. Bioelectron. 24(1), 1–13. 120. Edelstein, R. L., Tamanaha, C. R., Sheehan, P. E., Miller, M. M., Baselt, D. R., Whitman, L. J., and Colton, R. J. (2000). The BARC biosensor applied to the detection of biological warfare agents. Biosens. Bioelectron. 14(10–11), 805–813. 26 MICROBIOLOGICAL DETECTORS FOR FOOD SAFETY APPLICATIONS

121. Ruan, C. M., Zeng, K. F., Varghese, O. K., and Grimes, C. A. (2003). Magnetoelas- tic immunosensors: amplified mass immunosorbent assay for detection of Escherichia coli O157:H7. Anal. Chem. 75(23), 6494–6498. 122. Sandhu, A., Kumagai, Y., Lapicki, A., Sakamoto, S., Abe, M., and Handa, H. (2007). High efficiency Hall effect micro-biosensor platform for detection of magnetically labeled biomolecules. Biosens. Bioelectron. 22(9–10), 2115–2120. 123. Pal, S., and Alocilja, E. C. (2009). Electrically-active polyaniline coated magnetic (EAPM) nanoparticle as novel transducer in biosensor for detection of Bacillus anthracis spores in food samples. Biosens. Bioelectron. J. 24(5), 1437–1444. 124. Alam, J., Riaz, U., and Ahmad, S. (2007). Effect of ferrofluid concentration on electrical and magnetic properties of the Fe3O4/PANI nanocomposites. J. Magn. Magn. Mater. 314(2), 93–99. 125. Kryszewski, M., and Jeszka, J. K. (1998). Nanostructured conducting polymer composites - superparamagnetic particles in conducting polymers. Synth. Met. 94(1), 99–104. 126. Kim, J. H., Cho, J. H., Cha, G. S., Lee, C. W., Kim, H. B., and Paek, S. H. (2000) Biosens. Bioelectron. 14(12), 907–915. 127. Pal, S., Alocilja, E. C., and Downes, F. P. (2007). Nanowire labeled direct-charge transfer biosensor for detecting Bacillus species. Biosens. Bioelectron. J. 22, 2329–2336. 128. Pal, S., Setterington, E., and Alocilja, E. C. (2008a). Electrically-active magnetic nanoparticles for concentrating and detecting Bacillus anthracis spores in a direct-charge transfer biosensor. IEEE Sens. J. 8(6), 647–654. 129. Pal, S., Ying, W., Alocilja, E. C., and Downes, F. P. (2008b). Sensitivity and specificity performance of a direct-charge transfer biosensor for detecting Bacillus cereus in selected food matrices. Biosyst. Eng. 99(4), 461–468. 130. Tims, T. B., and Lim, D. V. (2004) J. Microbiol. Methods 59(1), 127–130. 131. Cheun, H. I., Makino, S. I., Watarai, M., Shirahata, T., Uchida, I., Takeshi, K. (2001). J. Appl. Microbiol. 91(3), 421–426. Voeller V05-c02.tex V1 - 12/04/2013 1:04pm Page 27

2 PROCESSING AND PACKAGING THAT PROTECTS THE FOOD SUPPLY AGAINST INTENTIONAL CONTAMINATION

Scott A. Morris University of Illinois at Urbana-Champaign, Urbana, Illinois

2.1 INTRODUCTION

Too often, the first reaction to a social problem is to attempt to find a technical solution, when technology cannot overcome social problems, only their means and circumstances. Although there are some processing and packaging steps that can be taken to indicate intentional contamination of food, it is not possible to add a simple, inexpensive component to existing systems to prevent a determined attack: real solutions are always imperfect, often more complex and usually more difficult. Quite apart from malicious human efforts, nature has been attempting to contaminate food products since the first drying and salting of grains, meats and vegetables provided for a longer-duration food supply, and most food processing operations have a culture of quality that is intrinsically designed to work against these threats. Many of the efforts in large-scale food contamination have been directed at detection of outbreaks of food poisoning in the population and then remediation after an outbreak occurs. The food industry, which is usually quite careful about quality and safety, already has coding and recall management practices in place. These have historically worked very well after problems are detected, but assume that the producer is acting in good faith; that the inspection, notification and recall systems operate as they are supposed to; and that the product itself is not counterfeit. Packaging, which is intrinsically designed to protect the product against many natural and man-made hazards, may protect against pilferage or low-level postprocessing contamination of products, but the most that can be achieved for many products at any practical cost and production level is an indication of tampering. Additionally, the requirements for global outsourcing of manufactured products, ingredients and components; global markets for finished goods, the persistent push to minimize the costs of ingredients and packaging systems; ceaseless just-in-time logistics systems that have replaced warehouses; and per- petual demands to maximize productivity impede many types of proactive contamination prevention.

27 Voeller V05-c02.tex V1 - 12/04/2013 1:04pm Page 28

28 PROCESSING AND PACKAGING

As a result of these and other factors that are discussed in this chapter, there is no magic gadget that can be added to the food processing packaging and distribution system to make it perfectly safe against intentional contamination. What can be done is to assess and manage risks responsibly and appropriately, implement detection steps and improvements in technology where needed to make contamination difﬁcult at all points in the food system, and to ensure that a response system that is capable of remediating problems on a timely basis is in place. This would represent a substantial improvement on the current system.

2.2 PROCESSING

Intentional contamination or destruction of the food supply, usually through “agroterror- ism” (malicious disruption at the crop level) has been a historic strategy for the disruption of populations and a long-standing fear during wartime. In a world increasingly occupied with asymmetric warfare on many different levels, the threat of a subtle toxin or custom-tailored organism pervading the food system is an increasingly viable threat. These agents may be introduced at some point in the manufacturing and distribution process, between harvesting the raw commodity and consuming the finished product, in order to reach a much larger percentage of the population with less chance of detection than with simple package tampering. Food processing systems are designed to produce a product that is safe and stable within its distribution environment. That environment might range from long-term shelf-stable foods such as cans, jars and Meals, Ready to Eat (MRE) rations for the military to shorter-duration products such as dairy products, bagged salads and refrigerated “fresh” pasta. Historically, processing has involved either altering the food to make it inhospitable to spoilage organisms with processes such as drying or pickling, or applying thermal sterilization (and moderate toxin denaturation in some cases) followed by containment in a hermetically sealed container that prevents recontamination. Newer preservation methods have involved controlling the temperature throughout the distribution cycle to retard growth, and alternative methods of sterilization and containment have been developed, but the principle remains essentially the same. The implementation of Hazard Analysis and Critical Control Point (HACCP) requirements for food processing plants has provided tools to find and manage vulnerabilities to naturally occurring hazards. HACCP can also provide optimal points for assaying for contaminants or inspection for disrupted seals or counterfeit goods, if analytical tools are available that can detect the agent used. Increased registration and security requirements for food processing plants have reduced access to the production facilities and the use of operational risk management (ORM) strategies taken from the aerospace industry (which faces critical dangers as a matter of course) have provided tools to help develop situationally appropriate safeguards [1]. Balancing this are high employee turnover rates and the difficulty of documenting workers, the broad range of ingredients from multiple sources that may be shipped without tamper indication or verification systems, as well as reliance on Certificates of Analysis for ingredient safety rather than verifiable in-house testing. Many of these factors leave the system open to attack. On a larger scale, processed food production is run on a “Just-in-Time” paradigm that distributes product as quickly as possible and makes “catching” contaminated products before sale very difficult if there is any delay between detection of contamination or Voeller V05-c02.tex V1 - 12/04/2013 1:04pm Page 29

PACKAGING 29 illnesses and the issuance of a recall and warnings. This delay may allow a contaminated product to be distributed and consumed by a broad segment of the population, turning a containable incident into a debacle. An example of this is discussed subsequently.

2.2.1 Counterfeit Products and Ingredients Counterfeit products in the United States precede the revolutionary war; one of the com- plaints Britain had against its colonies was that British containers were being refilled and resold with a variety of products of dubious quality. Indeed there is evidence that colonists imported empty bottles from Britain for the sole purpose of counterfeiting or mimicking British products [2]. Since then, most counterfeit products have been concentrated around objects of small physical size, difficulty of verification and very high value which maximizes the return/risk benefit for the counterfeiter. Counterfeit designer watches are much more lucrative that counterfeit potato chips and counterfeit pills are very easy to make and immensely profitable. Because of this, governmental anticounterfeiting efforts in the food, drug and cosmetic milieu have been prominently focused on the pharmaceutical industry due to the immediate harm done to the consumer, although the cosmetics industry faces a booming expansion in counterfeit, copycat and “parasite” goods, that are often severely contaminated (and often attract buyers who are not willing to pay for the “real thing” but are unaware of the risks) [3, 4]. Counterfeiting of drugs in the United States has become an item of considerable concern since diluted and nonsterile Procrit® and Epogen®, were held responsible for deaths and illnesses in 2002 and Lipitor® tablets in 2003 were found to be counterfeit [5–7]. This has been addressed by increased requirements for verification and distribution traceability, something that the pharmaceutical industry had avoided on a cost basis for some time but is finally coming into practice with bar codes and other technical additions in Europe and elsewhere [8]. Counterfeiting (or product swapping) of foods items has long been a problem with luxury items such as high-value spirits, wines and foods. Counterfeiting of more general foods is less likely unless there is a combination of high financial or tactical value and ease of manufacturing counterfeits [9]. One of the food categories that shows an ongoing problem with counterfeiting is seafood, of which approximately 80% is imported into the US. Because seafood is difficult to identify after processing, it has been misrepresented for years and can be deliberately mislabeled as a high-value species to increase profitability. This fraudulent mislabeling carries the risk of illness or death from both intrinsically toxic species, and species that have absorbed dietary toxins. In 2007, a seafood importer was found to have mislabeled seafood containing puffer fish (which carry tetrodotoxin, a potent neurotoxin with no antidote) as monkfish [10]. Subsequent congressional inquiries highlighted the low rate and poor coordination of safety or security inspection in seafood imports [11].

2.3 PACKAGING

Packaging plays three general primary functions in modern consumer usage; protection, utilization and communication. While the protection function is often thought of as protecting a product against damage or contamination, in the case of a particularly dangerous material (nuclear fuel rods, for instance) the primary purpose is just Voeller V05-c02.tex V1 - 12/04/2013 1:04pm Page 30

30 PROCESSING AND PACKAGING the reverse—protection of the general environment against the product itself. Beyond that, the most fundamental function of nearly any type of food packaging is to protect the product against postprocessing contamination and quality loss. Since Mother Nature is constantly providing clever and relentless threats to degrade or contaminate products, packaging has always had a role to play in this regard, and the intentional human element is a very small component of the challenges that most food packaging resists on a continuing basis. From a security standpoint, packaging can thus be thought of as implicitly resisting intentional contamination to a large degree, but with the added utility of potentially being able to indicate when intentional tampering has occurred, either as an intrinsic feature of the design or as a result of an added component or design feature. This capability is usually balanced against the perception of packaging as an expense to be minimized during high speed production, and the requirement that the indicator must be both robust and accurate. Packaging also has the capability of communicating, most often using label copy or other printed material but other communication measures may be used as well. Part of a security system for particularly vulnerable assets may thus involve communicating a veriﬁcation code, as well as displaying an intact tamper-evident device, although these approaches have many weaknesses as well. Finally, packaging must have some utility. It is possible to create an impenetra- ble package for food products, but they would be prohibitively expensive, difﬁcult to mass-produce on the astronomical scale required for some food packages (the United States alone consumes more than a quarter-trillion packaged soft drinks per year), and would defeat use by the consumer.

2.3.1 Packaging and Safety Assurance For packaging systems, the assurance of a product’s integrity and safety is involved in several roles—borrowed from cryptography—that can be described as authentication, integrity and non-repudiation [12]. The first of these, authentication, is the assurance that the product is that which is described on the label and not a counterfeit from another producer. As is discussed in more detail, this is the more likely scenario for a broad-scale breach of food integrity at the ingredient level, since it is both easier to implement and more effective in disrupting entire segments of the food supply. The second, integrity, assures the customer that the product in the package has not been modified after production whether by the substitution of other goods or contamination of the existing product. There are historical precedents for this, the most notorious being the contamination of Tylenol in the 1980s that led to stricter requirements for over-the-counter (OTC) drug packaging and began the process of tamper-indicating packaging components. Although this particular method of disrupting consumer confidence in a product may be somewhat effective and is very expensive to the product manufacturer, it is less effective at inflicting actual widespread harm or disrupting sales of a broad class of products since tampering is typically tightly confined to a single geographic region or product. The final function, non-repudiation, ensures that the original manufacturer cannot deny producing the product (in effect, verifying that they have produced it). Non-repudiation is seldom considered in manufacture of physical items, but warrants discussion since it bears directly on the issue of plausible deniability by a manufacturer that the product Voeller V05-c02.tex V1 - 12/04/2013 1:04pm Page 31

PACKAGING 31 is not something that they are responsible for, and are therefore absolved both from the responsibility for its production and from any liability from the harm it caused. In a litigious economy that is increasingly outsourcing its consumer goods and many kinds of food from a broad multitude of globally distributed manufacturing operations and contract operations, this issue may prove increasingly important.

2.3.2 Requirements for Tamper Evidence The ﬁrst tamper-evident requirements were applied to OTC drugs after the Tylenol tampering episode in 1982. Few, if any, of these devices provide substantial protection against even a marginally skillful person with modest resources; this is a favorite challenge for clever students who generally have little trouble with them. While other countries that have adopted tamper-evident packaging requirements have been much more speciﬁc in their material, structure and printing requirements [13], current regulations for OTC drugs in the United States simply demand that

“Each manufacturer and packer who packages an OTC drug product (except a dermato- logical, dentifrice, insulin, or lozenge product) for retail sale shall package the product in a tamper-evident package ... having one or more indicators or barriers to entry which, if breached or missing, can reasonably be expected to provide visible evidence to consumers that tampering has occurred [14].”

Many of the tamper-evident and “freshness” seals on foods and pharmaceuticals can be defeated quite easily, although they do provide indication of pilferage to consumers (and a very good seal for many products, at the expense of annoyance at having to remove them). Unfortunately, there is no legal requirement for tamper evidence or counterfeiting protection in the food supply chain beyond due diligence, good manufacturing practices, record-keeping and perhaps quality assurances for ingredients and components. There are only voluntary guidelines and similarly worded guidelines for food importers, dairy processors, and general food producers, processors and transporters to address this issue:

“inspecting incoming products and product returns for signs of tampering, contamination or damage (for example, abnormal powders, liquids, stains, or odors, evidence of resealing, compromised tamper-evident packaging) or “counterfeiting” (inappropriate or mismatched product identity, labeling, product lot coding or speciﬁcations, absence of tamper-evident packaging when the label contains a tamper-evident notice), when appropriate.” [15–17]

While these guidelines urge reliance on broadly based ORM strategies rather than sin- gular technical devices—an effective strategy when properly implemented—one of the most prominent weaknesses in the integrity of the food supply is the specter of unremedi- ated, ingredient-level contamination. Tampering with a single product may destroy sales of that particular product until the crisis is resolved, but tampering with a commodity-level ingredient can result in a broadly distributed incident that not only harms consumers but can destroy conﬁdence in the entire class of materials.

2.3.3 Tamper Indication Devices Some tamper indicators have been successfully used for many years. The vacuum in traditional canned foods has been seen to be an indication of both physical and microbial Voeller V05-c02.tex V1 - 12/04/2013 1:04pm Page 32

32 PROCESSING AND PACKAGING integrity, and consumers are almost universally aware that a swollen or damaged can or a jar with without a vacuum should be discarded. As packaging moves to lower cost and more material-efficient structures, the indications of integrity are much simpler; for example, an undamaged or unopened pouch or aseptic pack, but there is no secondary indicator (such as vacuum) to indicate integrity and the seals are easily duplicated or repaired after tampering. By contrast, a package that is intrinsically difficult to construct, fill and seal such as a modern soft drink can, provides a much higher degree of resistance to intentional contamination. The delicacy of the structure, pressurized contents and single-use open- ing would require the use of extremely specialized equipment and processes to duplicate effectively. While this is not completely unlikely, since everything from nuclear reac- tor parts to hundred dollar bills are counterfeited on a regular basis [18, 19], it raises the stakes considerably and drives serious tampering efforts toward a more efficient, upstream means of disrupting food supplies. Many of the most effective tamper evidence devices are integral parts of the package that must be visibly and irreversibly damaged to gain access to the contents of the package. The most annoying of these can actually restrict access to everyone—the theft-resistant clamshell packaging around small electronic devices is a good example—but they are effective. Some of these date back before the Tylenol incidents created a flurry of new devices, and were intended to prevent pilfering, or dilution of alcoholic beverages [20]. Newer developments are aimed at a broader market, but operate in a similar manner of permanently locking or fusing together during assembly and requiring obvious destruction to access the contents. For tamper evidence to really come into its own, it must be something other than an afterthought in a product component that is often regarded as an annoying expense to be minimized while choosing among standardized components and processes. Designed-in rather than added-on tamper evidence that requires irreversible disruption of the basic package structure to gain access to the product would be a considerable improvement for many stock package types and components.

2.3.4 Add-On Indicators Add-on indicators are usually shrink bands around the package closure, a tape seal that may include a holographic pattern, laminate, or other optical feature that is relatively hard to duplicate by a casual tamperer, or a seal under the main closure that is inductively fused to the container and must be removed before the product can be used. Hidden coding on the package or optical microparticulate taggants in the product itself may also be added to make tampering or counterfeiting more difﬁcult. This will deter casual pilfering or tampering but is unlikely to defeat the use of duplicate packages or other more complex attacks. Often these seal indicators can be “lifted”; removed and transferred to another package as well. Most of the add-on indicators are very low cost and require very few resources to defeat. Moreover, consumers are not readily aware of them unless they create a nuisance factor. Most consumers would presume that a ﬁrst squeeze of catsup that occurred without stopping to remove the seal to be a blessing. Further, since there are few requirements and no standardization, a case or shipment of bottles with a uniform type of seal added to the entire lot after contaminating the product would not be recognized as different or as unsafe. What is required is a move to higher design standards, already achieved in many industries from automobiles to electronics, where quality of tamper evidence is Voeller V05-c02.tex V1 - 12/04/2013 1:04pm Page 33

PACKAGING 33 designed in from the start rather than added as an afterthought. This, like the soda can, will require a much higher level of infrastructure to duplicate, and will therefore further deter the casual tamperer. Unfortunately, tampering indicators and particularly add-on tamper indication such as holographic films and patterned adhesive tapes suffer from a form of the “banknote paradox”. They need to be so well standardized that they are well known to the users (and can easily be spotted when something is amiss), but that standardization makes their counterfeiting more beneficial since there will be a wider range of products for them to be used on. This in turn drives the engraving and printing of banknotes into an ever-escalating spiral of elegant anticounterfeiting technology with counterfeiters often close behind (or in some cases ahead of) the legitimate users. For a single-use item that must be of minimal cost such as antitampering tape, bands or stickers—primarily added to demonstrate due diligence—the same kind of spiral could drive it out of existence. Manufacturers are already increasing the level of printing sophistication and therefore, cost in some types of seals. The main reason that this has not been a substantial problem to date is that there have been few store-level malicious tampering attacks. Also there is little direct benefit from counterfeiting the tamper evidence devices (unlike counterfeiting currency), and obviously damaged products produced by clumsy tamperers are not recognized as such and are quickly discarded at some point in the distribution chain as having been damaged in transit or during display.

2.3.5 Proactive Devices Although still experimental, work has been done to develop packaging structures and materials that will actively indicate other conditions that pose a threat to the safety of products. In the food industry, there is a great deal of interest in indication of microbial growth and temperature abuse. Temperature abuse, which may result in spoilage, must integrate time and ambient temperature exposure to create a thermal profile. This has been achieved relatively inexpensively with electronic devices, but most devices that have been marketed use a chemical reaction analog that attempts to mimic the heat transfer and thermodynamic properties of the product-package system. Other RFID-based devices have been developed to indicate tampering via RFID signal, or use GPS data to report or record distribution route disruption but are typically used on large, high-value products. On-package indicators that are accurate and cheap enough to be widely used are still being pursued and must meet an extraordinarily low price level. Similarly, work is being done on indicators that use binding-site chemistry to indicate specific spoilage organisms or toxins, but these suffer from the threat of false-positive indication as well as their own specificity; a broad spectrum detector would require a broad array of specific indicators and would be complex and prone to a high false-positive rate. Single-hazard detectors are finding potential markets in rapid detection testing for production systems to provide fast, accurate quality control and indication of contamination [21]. Having rapid detection methods will lower the cost and increase the accuracy of screening for some hazards to the point where even small manufacturers can have a good degree of product safety assurance based on their own data, rather than assurances from suppliers that deflect liability. This would be an enormous “grass roots-level” improvement for detection, interruption and containment of contamination episodes. Voeller V05-c02.tex V1 - 12/04/2013 1:04pm Page 34

34 PROCESSING AND PACKAGING

2.3.6 Optical Systems Since many of the systems rely on customer inspection, specialized readers may not be available, but there is a wide range of sealing materials and devices that can aid in authentication as well as tamper detection if properly constructed. These usually depend on specialized printing processes and materials that require technical sophistication to produce, which will reduce the likelihood of casual counterfeiting and tampering, but as with the $100 “supernotes”, are not impossible to duplicate given sufficient resources. Indeed, many of the “speciality” features such as color-shifting inks and fracture-evident dyes can be approximated with materials purchased in art supply stores. Holographic images, microprinting, frequency-specific pigments or additives that react to certain wavelengths of light and may change when damaged, and retroreflective or interferometric imagery that may contain both overt and covert features—often requiring a specialized viewer to resolve—can reduce counterfeiting or alteration. Many similar features are used to authenticate drivers’ licenses.

2.3.7 Physical “Token” Systems and RFID Wax and clay seals date into prehistory as a means of guaranteeing the authenticity of documents and products, and the inclusion of a physical veriﬁcation token that is difﬁcult to reproduce might offer some protection against counterfeit products. Software from major manufacturers has often been shipped with a complex authenticity seal that includes activation codes useable only with that copy. RFID devices carrying an authentication code encrypted in memory also have been proposed but neither of these systems, nor most systems relying on a physical object, eliminate the problems of “lifting”—removal of the token or indicator from one product to use in another—very well, and are still too expensive for individual consumer food package use. Additionally, some simpler RFID systems are quite easily cloned or have their encryption broken, even at a distance, and are the subject of a great deal of controversy because of their widespread use in passports, bank cards and other devices [22–24].

2.3.8 Product Authentication For products such as pharmaceuticals that are increasingly targets of counterfeiters and potentially targets for attack, authentication systems have begun to be implemented. For food products, most of which already carry batch coding of some type, there is less impetus for these both because of complexity and cost. These systems typically depend on one or several systems for authentication which may include bar codes, special printing features such as moire´ images and light-frequency-specific inks and product taggants that require a reader to translate, or numeric codes that can be verified via websites. A more intriguing possibility lies with the potential to trap contaminated products during distribution or at the final point of sale by interfacing with inventory or point of purchase data systems to flag products as they are being shipped, shelved, and checked out of the store. This would require that current Universal Product Code (UPC) coding schemes include batch number information, but would also return low-cost data about batch shelf-time and turnover of product, something that still is often tediously hand-collected by manufacturer’s representatives in stores. Voeller V05-c02.tex V1 - 12/04/2013 1:04pm Page 35

PACKAGING 35

2.3.9 Multipart Authentication Many secure systems already require multipart identification and authentication. The combination of usernames and passwords on e-mail accounts is a good example. Sim- ilarly, the addition of secure authentication methods for encrypted data transfer systems such as PayPal® and others have made electronic commerce possible. Modern combina- tions of identification and authentication have made the use of private information fairly secure. For packaged goods, it may be possible to utilize a matched coding system where customers or retailers can authenticate code numbers against batch numbers to verify their authenticity [25]. Similar systems have been used to activate software, cell phones and credit cards for some time. In theory, this could be almost entirely automated and only require a web page lookup for authentication, although this might present other vulnerabilities (website hacking or posting a false authentication webpage and then printing the counterfeit product with the false webpage’s address to fool the consumer). The addition of additional features (such as product type, size, count that should be present in the validated product) can add another layer of security since a mismatch of product type or size even with an authenticated code would immediately raise suspicions. The problem with these types of systems that may be quite useful with relatively low-volume items like pharmaceuticals, is the sheer quantity of material that is handled. Very few busy people will spend time authenticating a week’s worth of groceries for a large family, and restaurant operations would be even more hectic than they are already. Instead, most consumers (if they consider it at all) depend on the historical use of media dissemination of information about product recalls based on lot numbers or code systems already used by manufacturers. While this is useful to consumers when problems are spotted and disseminated properly, it does not help to prevent the initial outbreak often required to flag the problem along with the attendant illnesses.

2.3.10 Security and the Base Rate Fallacy For a security authentication or tamper evidence system to be acceptable, it must both very reliably detect attacks and even more reliably reject “false alarms” and it is usually the latter condition that is hardest to satisfy. Much as a smoke alarm that goes off at the slightest provocation is quickly disabled, safety indicators or systems that are constantly inaccu- rate are quickly disabled, discarded or ignored. Given the billions of units of packaged food handled daily, even an improbably small percentage of false alarms can cause large numbers of people to ignore the indicators or, worse, to panic over many perfectly good products being ﬂagged. For example; if a detection system has a 0.1% false-positive rate, and the rate of actual occurrence (tampering or contamination) is one in a billion units, surveying a billion units will trigger 1,000,001 alarms (one “true” alarm plus 1,000,000 “false” alarms) and will still only give a one in a million chance of being correct. For this reason, nearly all attempts at safety measures to ensure rigorous 100% inspection and validation in food manufacturing fail, and indeed this is the reason that many medical tests are repeated after a serious condition is diagnosed [26]. Risk management therefore becomes more a matter of building security into the system beforehand using ORM principles, rather than relying solely on inspecting it after the product is completed, a lesson taken from the “Total Quality Management” (TQM) principles that transformed Japanese manufacturing into exemplars of production quality [27]. Voeller V05-c02.tex V1 - 12/04/2013 1:04pm Page 36

36 PROCESSING AND PACKAGING

2.4 SYSTEM FLEXIBILITY AND RESPONSE

While postprocess contamination is an ongoing concern, the realities of the tactical effectiveness of a contamination, package tampering or counterfeiting episode must be considered in a larger context of both, effectiveness in disrupting confidence in the food supply and ease of implementation. For an attacker to be effective these are both highly desirable results and a food safety system must likewise work to minimize these benefits. As has previously been discussed, the ability of the existing food safety system to respond in an appropriate and timely manner is a primary component of any food system’s security. While it may be minimally possible to shut down an airport in the event of a person bypassing security screening, it would be impossible to shut down the entire food production system. Additionally, few food customers would endure the kind of security screening required at airports to buy a liter of milk on their way home from work. Thus, access control for the food system is only feasible at the production and transportation level, although communication and indicators can be used at the consumer level. Most retail-level food tampering is either accidental damage or the result of pilfering, often the result of putting candy and cereals on low shelves that children can easily reach. These and other “naturally” spoiled products are usually discarded or held for replacement on discovery by the store staff unless there is an obvious, recurring problem that requires contact with the manufacturer. This provides very little systemic information to guess where failed attempts at malicious tampering may have occurred. It is difficult to determine under the best of circumstances which of those problems may have been caused by shipping damage, in-store damage or pilfering, or might have been the result of an attempt at malicious tampering. It is similarly unrealistic to expect every food retailer to accurately diagnose and report problems with every unsaleable product they discard. Thus we are left with examining those cases which are clearly the result of contamination or tampering, unfortunately too often involving incidents of injury, illness or death. Once the determination of a contamination or tampering outbreak has occurred, the resilience and responsiveness of the system is an utterly critical part of providing a timely response that does not shut down the entire national food supply. Because of this, good security systems avoid “brittleness” that is, abrupt shutdown of the asset they are designated to protect when an incident occurs and to fail to or isolate the problem in a “resilient” fashion (i.e. a timely, useful and resource-efficient manner) [28]. Designing these responses in all types of industries is an ongoing, adaptive process (and is somewhat predictive in the best situations) that seeks to minimize the contamination or failure “space” after an incident while maintaining as much of the surrounding network of supply as possible [29]. Unfortunately, incidents that have occurred highlight the fragmented, opaque, uncoordinated, and unresponsive nature of the food safety system which makes it highly vulnerable to a systemic attack even using very simple methods.

2.4.1 Cascading Failure in the Food Processing and Packaging System Cascading failure is generally deﬁned as a failure that triggers a succession of down- stream failures in other elements in a system of some type, such as a failure at a single electrical switching station that triggers a more massive power outage [30]. The food industry processes and combines a myriad of both naturally occurring and synthetic materials to provide our modern diet and this provides a target-rich environment for intentional contamination. Fortunately, nature has preceded this by providing a huge Voeller V05-c02.tex V1 - 12/04/2013 1:04pm Page 37

SYSTEM FLEXIBILITY AND RESPONSE 37 range of naturally occurring hazards ranging in scale from oxidation to vermin. The food processing industry has had mechanisms in place for many years to inspect critical points during production and then initiate recalls in the event of accidental misprocessing or contamination. It generally works quite well if used in a timely manner. However, for a response system to be effective it must be used and it must actually respond. Beginning before the events of 9/11, the Government Accounting Office has repeatedly highlighted the fragmented, uncoordinated and resource-poor nature of the food safety system with regard to intentional contamination [31–34]. Additionally, the opacity of the system contributes to delay and inaction. Under current regulations, disclosure of internal safety audits to government officials are not required, and recalls must be requested from company officials, slowing notification and adding delay during outbreaks of contamination [35]. This delay and restriction slows the response of the food safety system, adding to the severity of outbreaks and threatening whole commodity sectors when simple product recalls might have sufficed. No simple change or device for the processing or packaging technology can overcome this. A culture of quality and safety improvement based on improved practices, technology and regulation will have the biggest impact.

2.4.2 Safety System Failure Case Study: Peanut Corporation of America The incidents of Peanut Corporation of America (PCA) intentionally shipping salmonella-contaminated peanut butter ingredients to other packers and food manufacturers showed the infrastructural weaknesses that may occur when safety measures break down. Inspectors of the plant failed to change practices at the facility, despite a similar incident beginning in 2004, in a similar processing plant owned by Con Agra only 75 miles away that had gone uncorrected for three years after notification of the FDA by a whistleblower, and subsequent company refusal to release test results. Con Agra’s problems were only addressed when illnesses were finally reported, and even then their test records were never made public [36]. Although Con Agra completely rebuilt and improved its facility and operating procedures, PCA continued to operate with apparent impunity until deaths and illnesses were reported in 2008–2009, and the failure cascaded outward [37, 38]. A second PCA plant in Texas that was operated producing other nut products was never inspected at all, since it was never registered with the Texas Department of Agriculture although it shipped products nationally. As health problems came to light, other food manufacturers that used ingredients supplied by PCA were abruptly forced to recall nearly 4000 distinct products at tremendous expense, depressing peanut product sales by nearly 25% [39–41]. Many of these products had depended on the supplier’s assurance—either informally or with Certificates of Analysis—to ensure the safety of their ingredients (or at least the absolution of liability) since many small operations do not have the capacity to do safety analyses. Thus, a supplier who ships contaminated product as an ingredient that is subsequently used in a plethora of other products highlights both the complexity and vulnerability of the food manufacturing system, and how a brittle failure could cripple an entire commodity sector by destroying public confidence. From this we can see how an intentional incident could easily be caused by simple contamination or counterfeiting of a commonly used ingredient that is distributed through a system where the safety measures have broken down or are being ignored. Since food ingredients are not subject to even the minimal regulatory requirements of tamper Voeller V05-c02.tex V1 - 12/04/2013 1:04pm Page 38

38 PROCESSING AND PACKAGING evidence applied to other products, and since there is often very little distinctive printing or packaging involved, the potential for contamination and counterfeiting are enormous. Often the only motivation for ingredient manufacturers to supply validation or tamper evidence features is the threat of involvement in legal action, and so as with consumer products, they tend toward the minimums necessary to prove due diligence on the part of the supplier. This may be a numbered tag or seal that must match the invoice that may have been sent separately by e-mail or fax; an excellent start, but hardly proof against a diligent tamperer.

2.5 CONCLUSION

While there are many types of technology that are being developed to make the food production and packaging system more secure, most of these fail in some aspect of reliability, simplicity, dependability, or consumer recognition. To balance this, the food processing, packaging and distribution industries have been fighting a pitched battle against numerous naturally occurring hazards on an ongoing basis for centuries. This has provided some remedial mechanisms for dealing with contamination incidents once they are detected and reported. The most useful practical approach is to assess threats accurately and manage risks appropriately, implement practical and useable detection methods where needed to make contamination difficult at all points in the food system, and to ensure that a response system that is both capable and operational is in place to contain and remediate intentional incidents on a timely basis. An ORM strategy may provide a very good measure of prevention against casual tampering, but will likely not prevent a carefully targeted attack, particularly from intentionally contaminated ingredients that are created at a high level and widely distributed to product manufacturers. On-package detectors and tamper indication can “keep honest people honest” and give some indication of contamination, but are subject to all kinds of errors and lack of consumer awareness. There is a need for better design and a higher degree of upstream integration of tamper evidence in the package design process, rather than simply discharging fiduciary duty with adhesive tape and shrink wrap. Detection methodologies, while being developed for rapid detection in plant operations and other inspections, may never pass into in-package use because of the combination of cost, possible toxicity, specificity of test and unacceptable false-positive rates. These can be extremely useful in the detection of contamination during spot checks as ingredients move from supplier to production lines. This is a function that has been abandoned in many operations in favor of “paper” assurances that avoid liability, creating enormous security loopholes; besides, small-manufacturer testing could be an enormously lucrative market for testing-kit manufacturers. Validation of products via multipath (package coding plus lookup lists on-line) might allow retail checkout or consumer screening of products once alerts have been sounded, but are unlikely to be used on a persistent basis by many consumers in the course of their day-to-day lives except as a spot check in the event of well-publicized recalls. Finally, remediation systems exist, and can work quite well if used properly. A well-publicized recall can remove products from circulation very quickly, though perhaps at the cost of some good product being discarded. For a broad class of alerts such as the peanut butter outbreaks discussed in this chapter, it could also cause a depression in the sales for that particular item, but these are typically short-lived phenomena and Voeller V05-c02.tex V1 - 12/04/2013 1:04pm Page 39

REFERENCES 39 might be considered an acceptable and insurable security cost. Most importantly, if the detection-remediation system continues to be opaque, failing to act or delaying action for a very long time as it has done very badly in the given case and others, an outbreak may proliferate for months or even years until the symptoms, illnesses or deaths accumulate to a signiﬁcant level, and the massive response is cripplingly brittle rather than adaptive. As has been too-adequately demonstrated, there is little difference in method between a malicious manufacturer operating for proﬁt and a malicious fabricator trying to generate the public fear that is the hallmark of terrorism.

REFERENCES

1. Federal Aviation Administration (2008). FAA Systems Safety Handbook. Chapters 3,15 and Appendices e and h are recommended. Available at http://www.faa.gov/library/manuals/ aviation/risk management/ss handbook/. 2. Hine, T. (1995). The Total Package: The Evolution and Secret Meanings of Boxes, Bottles, Cans, and Tubes. Little, Brown, & Co., New York, pp. 49–50. 3. (November 28, 2003). Counterfeit Cosmetics Challenge Market Cosmetics Design Maga- zine. Available at http://www.cosmeticsdesign.com/Products-Markets/Counterfeit-cosmetics- challenge-market. 4. (October 12, 2007). The Truth About Fake Beauty. Available at http://www.beautyheaven. com.au/article/perfume-the-foul-scent-of-a-fake. 5. U.S. Food and Drug Administration (2003). Counterfeit Procrit. Available at http://www. fda.gov/ForConsumers/ByAudience/ForPatientAdvocates/HIVandAIDSActivities/ucm125089. htm. 6. U.S. Food and Drug Administration (2002). Important information about counterfeit EPOGEN . Available at http://www.fda.gov/ForConsumers/ByAudience/ForPatientAdvocates/ HIVandAIDSActivities/ucm125109.htm. 7. U.S. Food and Drug Administration (2005). Counterfeit Drugs” Statement of Randall W. Lutter, Ph.D., Acting Associate Commissioner for Policy and Planning Food and Drug Admin- istration before the Senate Subcommittee on Criminal Justice, Drug Policy, and Human Resources, Committee on Government Reform, November 1 . Available at http://www.fda.gov/ NewsEvents/Testimony/ucm112670.htm. 8. Taylor, P. (2008). EFPIA Says Traceability Pilot Will Start Next Year. Available at http://www. outsourcing-pharma.com/On-your-radar/Patient-safety/EFPIA-says-traceability-pilot-will-start- next-year. 9. Bertrand, K. (2009). Improve Security Through Packaging. Food Processing February 14, 2009 . Available at http://www.foodprocessing.com/articles/2006/027.html. 10. US Food and Drug Administration (2007). FDA Warning on Mislabeled Monkﬁsh. Available at http://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/2007/ucm108920.htm. 11. Government Accounting Ofﬁce (2007). Seafood Fraud: FDA Program Changes and Better Col- laboration among Key Federal Agencies Could Improve Detection and Prevention. Available at http://www.gao.gov/new.items/d09258.pdf. 12. Schneier, B. (1996). Applied Cryptography. John Wiley & Sons, New York, pp. 1–3. 13. (2003). Code of Practice for the Tamper Evident Packaging (TEP) of Theraputic Goods, 1st ed.. Australian Government Department on Health and Ageing, Therapeutic Goods Administration. Available at http://www.tga.gov.au/DOCS/pdf/tepcop.pdf. 14. 21 CFR 211.132 Current Good Manufacturing Practices for Finished Pharmaceuticals. Avail- able at http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?fr=211.132. Voeller V05-c02.tex V1 - 12/04/2013 1:04pm Page 40

40 PROCESSING AND PACKAGING

15. FDA (2007). Guidance for Industry. Importers and Filers: Food Security Preventative Measures Guidance. Available at http://www.fda.gov/Food/FoodDefense/FoodSecurity/default.htm. 16. FDA (2007). Guidance for Industry, Food Producers, Processors, and Transporters: Food Security Preventive Measures Guidance. Available at http://www.fda.gov/Food/Guidance ComplianceRegulatoryInformation/GuidanceDocuments/FoodDefenseandEmergencyResponse/ ucm083075.htm. 17. FDA (2007). Dairy Farms, Bulk Milk Transporters, Bulk Milk Transfer Stations and Fluid Milk Processors: Food Security Preventive Measures Guidance. Available at http://www.fda.gov/ Food/GuidanceComplianceRegulatoryInformation/GuidanceDocuments/FoodDefenseand EmergencyResponse/ucm083049.htm. 18. EPRI Nuclear Executive Update (March, 2009). Growing Risk of Counterfeit Items Emphasizes Need for Industry Guidance. Available at http://mydocs.epri.com/docs/CorporateDocuments/ Newsletters/NUC/2009-03/3a.html. 19. Mihm, S. (2006). No Ordinary Counterfeit. New York Times July 23. Available at http://www. nytimes.com/2006/07/23/magazine/23counterfeit.html? r=1&scp=1&sq=No%20Ordinary %20Counterfeit&st=cse. 20. (1913). U.S. Patent 1,065,211. Bottle-Stopper. www.uspto.gov. 21. FDA (2003). Testing for the Rapid Detection of Adulteration of Food, Report to Congress. Available at http://webharvest.gov/peth04/20041021081954/www.fda.gov/oc/bioterrorism/ report congress.html. 22. Hacker War Drives San Francisco RFID Cloning Passports. Engadget, 2/2/09 . Available at http://www.engadget.com/2009/02/02/video-hacker-war-drives-san-francisco-cloning-rfid- passports/. 23. Chris Paget’s talk on drive-by RFID cloning. Available at http://video.google.com/videoplay? docid=-282861825889939203, with slides Available at uploads/paget shmoocon edl.ppt. 24. (2005). RFID Chips in Car Keys and Gas Pump Pay Tags Carry Security Risks. Press Release Johns Hopkins University. Available at http://www.jhu.edu/news info/news/home05/jan05/ rfid.html. 25. Johnston, R. G. (2004). LAUR-04-8055 An Anti-Counterfeiting Strategy Using Numeric Tokens. Available at http://www.fda.gov/OHRMS/DOCKETS/DOCKETS/05n0510/05N-0510-EC4- Attach-1.pdf. 26. Elmore, J. G., et al. (1998). Ten-year risk of false positive screening mammograms and clinical breast examinations. N.Engl.J.Med.338(16), 1089–1096. Available at http://www.ncbi. nlm.nih.gov/pubmed/9545356. 27. BPIR.com History of Quality. Available at http://www.bpir.com/total-quality-management- history-of-tqm-and-business-excellence-bpir.com.html. 28. Haimes, Y. Y., et al. (2008). Homeland security preparedness: balancing protection with resilience in emergent systems. Syst. Eng. 11(4), P287–P308. 29. Beal, J. (2003) Near-Optimal Distributed Failure Circumscription AI Memo 2003-017. Massachusetts Institute of Technology Artificial Intelligence Laboratory. Available at ftp://publications.ai.mit.edu/ai-publications/2003/AIM-2003-017.pdf. 30. Dobson, I., et al. (2007). Complex Systems Analysis of Series of Blackouts: Cascading Fail- ure, Critical Points, and Self Organization. Chaos 17, 026103 (1–12). American Institute of Physics. Available at http://eceserv0.ece.wisc.edu/∼dobson/PAPERS/dobsonCHAOS07.pdf. 31. U.S. Government Accounting Office (1999). Food Safety: Agencies Should Further Test Plans for Responding to Deliberate Contamination. Available at http://www.gao.gov/new.items/ rc00003.pdf. 32. U.S. Government Accounting Office (2003). Voluntary Efforts are Under Way, but Federal Agencies Cannot Fully Assess their Implementation. Available at http://www.gao.gov/new. items/d03342.pdf. Voeller V05-c02.tex V1 - 12/04/2013 1:04pm Page 41

REFERENCES 41

33. U.S. Government Accounting Ofﬁce (2007). Federal Oversight of Food Safety: High-Risk Designation Can Bring Needed Attention to Fragmented System. Available at http://www. gao.gov/new.items/d07449t.pdf. 34. U.S. Government Accounting Ofﬁce (2008). Federal Oversight of Food Safety: FDA has Pro- vided Few Details on the Resources and Strategies Needed to Implement its Food Protection Plan. Available at http://www.gao.gov/new.items/d08909t.pdf. 35. Harris, G. (2009). Peanut Product Recall Took Company Approval. New York Times. February 3, 2009 . Available at http://www.nytimes.com/2009/02/03/health/policy/03peanut.html. 36. Moss, M. (2009). Peanut Case Shows Holes in Safety Net. New York Times, February 9, 2009 . Available at http://www.nytimes.com/2009/02/09/us/09peanuts.html. 37. Harris, G. (2009). Peanut Products Sent out Before Tests. New York Times, February 12, 2009 . Available at http://www.nytimes.com/2009/02/12/health/policy/12peanut.html. 38. Centers for Disease Control (2009). Investigation Update: Outbreak of Salmonella typhimurium Infections, 2008–2009. Update for April 29, 2009 (FINAL web update). Available at http://www.cdc.gov/salmonella/typhimurium/update.html 39. Cook, C. (2009). Peanut Recall’s Ripples Feel Like Tidal Wave for Some Companies. New York Times February 26 . Available at http://www.nytimes.com/2009/02/26/business/ smallbusiness/26sbiz.html. 40. Martin, A., and Robbins, L. (2009). Fallout Widens as Buyers Shun Peanut Butter. New York Times February 7th. Available at http://www.nytimes.com/2009/02/07/business/07peanut.html. 41. Harris, G. (2009). Peanut Recall Leads to Criminal Investigation. New York Times January 31. Available at http://www.nytimes.com/2009/01/31/health/31peanut.html? r=1. Voeller V05-c02.tex V1 - 12/04/2013 1:04pm Page 42 3 EARLY DETECTION AND DIAGNOSIS OF HIGH-CONSEQUENCE PLANT PESTS IN THE UNITED STATES

Kitty F. Cardwell and William J. Hoffman United States Department of Agriculture, Washington, D.C.

3.1 INTRODUCTION

There exists well-documented historic precedent of the intentional use of biological organisms as weapons to strike against a target enemy either directly (human pathogens) or by adversely impacting its agricultural security (animal and plant pathogens) [1–3]. Gov- ernment sponsored research into the development of biological weapons for use against humans, livestock, and crops was prevalent during the early decades of the twentieth century. Most government bioweapons programs included research on the culture and testing of disease agents intended speciﬁcally for use against livestock and food crops. There is concern because no elaborate delivery technologies or methods are necessary for clandestine, economically targeted bioweapon attacks on agricultural crops [2]. Many exotic plant pathogens are already highly infectious with high reproductive potential that facilitate rapid exponential epidemic development when environmental conditions are favorable [3, 4]. Furthermore, plant pathogens are generally not infectious to human han- dlers; inocula are available from infected crops around the world; and collection, increase, and delivery to a target crop and region is not technologically difﬁcult. The US National Research Council concluded in a 2002 report that the potential and (consequences of) deliberate bioterrorism attacks directed at US agriculture needs to be recognized as a serious threat to the United States and its agricultural economy [3, 5]. Bioterrorism is perhaps the most extreme example of the larger issue of invasive species of exotic pests and pathogens to new areas as a result of deliberate or inadvertent human activity or of more ”natural” spread [2]. Agricultural emergencies, whether deliberately caused or natural, follow the same notional phases as any disaster, including biological and chemical attacks on agriculture, animals, or civilians. Phase 1 includes detection and diagnosis; phase 2 involves a systematic monitoring to characterize and delimit the area of outbreak and mitigation in

43 44 DETECTION AND DIAGNOSIS OF PLANT PESTS

the outbreak area; and phase 3 is the longer-term response and recovery activities. This chapter concerns phase 1—early detection and diagnosis of a new biological threat to agriculture or valuable natural resources [6]. In the United States, there are an estimated 1 billion acres of crop, forest, and range lands. It is physically impossible to closely monitor all of this area to achieve early detection of newly introduced pests or pathogens. As such, these resources make strategically and tactically attractive targets to those who would strive to provoke food shortages, loss of valued ecosystem, economic damage either through loss of trade or loss of com- petitiveness, public loss of conﬁdence in food safety and security, or direct damage to humans and/or animals [3]. Agricultural crops are particularly vulnerable due to ease of access and to the logistical challenges of continuous surveillance. In 2004, an estimated 155 million acres were planted with corn and soybeans, alone. Numerous other crops are grown over large acreages and/or are high-dollar intensively grown specialty crops such as grapes, citrus, and vegetables. Because of the challenges of surveillance over such a vast area, there is the potential of a long lag time between the introduction of a pathogen or pest (intentional or natural) and the detection. Likewise, any delay in diagnostic processing once detected, would further delay appropriate response. For example, citrus canker, a bacterial disease of oranges, is believed to have been in Florida at least 2 years before it was detected and diagnosed [7]. Likewise, the emerald ash borer, a devastating invasive insect pest of ash trees, was estimated to have been in Michigan 5 years before it was discovered near Detroit, Michigan [8]. During this time, these pests and diseases spread to the extent that there is little hope that containment and eradication can be successful. Once an exotic or invasive organism has been detected, the need for intensive monitoring and sample diagnosis during a phase 2, also creates signiﬁcant logistical challenges. Successful execution of a monitoring, early detection, and diagnosis program improves the chances of responding quickly, minimizing impact, and reducing time to total recovery. The purpose of this chapter is to explore the technical and infrastructural challenges to early detection, and rapid and accurate diagnosis of high-consequence exotic pests and pathogens.

3.2 DETECTION: MONITORING AND SURVEILLANCE

High-consequence pest monitoring and surveillance is a series of activities that combines biological science and human performance. This section will examine three different sub- sets of these efforts: formal human (active) systematic surveillance, nonformal human (passive) surveillance, and remote or automated surveillance. The terms formal and nonformal, indicate the level of government oversight and organization. These activities will be analyzed using a variation of Gilbert’s behavioral engineering model [9]; examining the capacity, knowledge, and motivation of the individuals and organizations involved.

3.2.1 Formal (Active) Human Surveillance Formal human active surveillance of high-consequence exotic crop pests are led by the US Department of Homeland Security (DHS)’s Customs and Border Inspection (CBI) Program and the United States Department of Agriculture (USDA)’s Animal and Plant Health Inspection Service (APHIS). These organizations work together in surveillance DETECTION: MONITORING AND SURVEILLANCE 45 attempts designed to exclude pests from coming into the country. APHIS works to monitor regulatory pest problems for domestic establishment [10].

3.2.1.1 Exclusion. The homeland security act of 2002 transferred federal agricultural port inspection responsibility from APHIS’ Plant Protection and Quarantine (PPQ) division to the DHS-CBI [11]. The bulk of these activities to exclude 6000 miles of borders and 100,000 miles of shoreline [12] take place at 317 border inspection stations, 161 of which are staffed with agricultural specialists [11]. The daily trafﬁc across these borders includes 1.1 million people, 45,000 trucks, 550 vessels, 2500 aircraft, and 341,000 personal vehicles [12]. Each day, an average of 1145 food product seizures take place as a result of short range electronic surveillance, interviews, canine sweeps, and hand searches. APHIS conducts an extensive training program and procedure manual development, provides guidance on inspection targeting and alert notiﬁcation, and collaborates on port of entry review in order to bolster Customs and Boarder Patrol (CBP) efforts [11]; and thus remains involved with border pest exclusion.

Capacity. While efforts of the CBP agricultural specialists are laudable, there are two important reasons why they cannot possibly exclude all high-consequence pests and pathogens from entering the country. First, not all such pests enter the United States through registered border crossings. This includes pests that may enter the country naturally through the air or are carried by those crossing the border illegally [13]. Second, to intercept those pests and pathogens that do go through the inspection process, additional requirements would be needed to include fumigation of all manner of conveyances from countries where target pests are found as well as more intensive questioning and inspection of foreign visitors and returning citizens [14]. Knowledge. The CBP agricultural specialists are adequately trained [11] to carry out their mission. However, cross agency communication has impeded implementation of procedure manual updates and agricultural pest alerts. This has caused delays in getting timely information to those on agricultural homeland security’s front lines [11]. Motivation. Legislators, farm organizations, and the National Plant Board have argued that agricultural border protection would be better off with the USDA, a department more focused on pest exclusion and food security [15–17]. Agricultural pest exclusion is one of many important priorities of CBP, an organization focused on the broad mission of ensuring safety and security while facilitating trade and travel. Within CBP, agricultural related border security competes for resources in an extremely broad and high consequence threat pool. District CBP agricultural liaisons provide regular communication to ﬁeld managers, which helps pest exclusion receive the operational focus it deserves [11].

After a high-consequence crop pest does cross the US border, detection depends on domestic monitoring systems, both formal and nonformal surveillance activities.

3.2.1.2 Monitoring. The USDA estimates that introduced plant pests account for annual agricultural losses of $41 billion and it provides leadership for monitoring domestic and regulated pests. An important formal monitoring program is the Cooperative Agricultural 46 DETECTION AND DIAGNOSIS OF PLANT PESTS

Pest Survey (CAPS) operated by APHIS. The CAPS provides resources to state departments of agriculture in all 50 states and 3 US territories to track more than 400 pests [18]. Approximately 40 of these pests are identified as national priorities by the CAPS program and the balance are proposed as local priorities. Detected high-consequence crop pests that are subject to regulatory action are immediately reported by CAPS survey participants to APHIS/PPQ emergency programs staff and are sent for confirmation by that group’s national identification services. All survey data is submitted to the National Agricultural Pest Information System to contribute to national analysis. Soybean rust, a serious soybean pathogen, was detected late in 2004. Prior to the 2005 soybean growing season, the USDA developed a sentinel plot based surveillance system to monitor disease progress throughout the year and provide near real-time information for pest managers. This has developed into the integrated pest management Pest Infor- mation Platform for Extension and Education (ipmPIPE), a partnership funded by the USDA and client industry, and managed by the cooperative extension systems across the country [19].

Capacity. In 2006, the CAPS program distributed approximately $5 million to all of the states and US territories combined. The state departments of agriculture participants in the CAPS program have lamented that the cooperative agreements they receive can only fund one survey coordinator and a few survey programs [18]. Similarly, the ipmPIPE has yet to achieve a stable funding source for the $2.277 million needed for core operations. While these efforts are highly leveraged through state government and land grant university funds, this level of funding does not purchase the type of capacity needed to thoroughly and systematically monitor high-consequence pests. Knowledge. These programs thoughtfully set national priorities and the decentralized nature provides the ground truths from the states and regions. State departments of agriculture and state cooperative extension service’s often work hand in glove with land grant university researchers so that pest monitoring programs are informed by timely research [21]. Motivation. State Plant Regulatory Officers who participate in the CAPS program are the lead state government officials in PPQ issues. These individuals are highly motivated to monitor plant pests and correct problems before they are uncontrol- lable. Cooperative extension officials are motivated to collect data from an academic perspective and to act as an information hub in service to their clientele. This motivation is particularly keen when their clientele faces an imminent threat. However, focus on such a threat could draw attention away from other vulnerable pathways.

3.2.2 Nonformal (Passive) Human Surveillance There are many people involved with crop pest management who are not involved with government sponsored surveillance. Approximately 90% of row crops and vegetable acres are scouted for endemic weeds, insects, and/or diseases that routinely threaten their crop’s profitability. The vast majority of this scouting is performed by owner operators. Approximately 10% of the row crop acreage is scouted by chemical dealers and a similar amount is scouted by independent crop consultants. Certified Crop Advisor (CCA) utilization increases significantly for vegetable and other high-value crops [22]. ESTIMATING RISK TO ORIENT SURVEILLANCE 47

During the last 5 years, cooperative extension services have conducted a campaign to incorporate exotic and invasive weed, insect, and disease detection into normal scouting activities. Training “first detectors”, an activity led in part by the National Plant Diag- nostic Network (NPDN), includes instruction on (i) the importance of high-consequence exotic plant pest detection; (ii) what scouts should do if they see something they do not recognize; and (iii) exotic pests that crops scouts should be on the lookout for in addition to routine pests. Many training participants were added to a first detector “registry” that will allow the NPDN to alert them based on recent pest finds and conditions favorable to disease presence [23]. Timely communications from the NPDN, coupled with information from the ipmPIPE, farm press county extension agents, and other information sources, could help crop scouts keep alert for exotic pests as they perform regular scouting activities.

Capacity. In addition to the operators of the 2 million US farms [24], there are 14,000 certified crop advisors [25] who make regular continuing education a part of their professional development. These crop advisors include chemical manufacturer representatives, chemical retailers, extension agents, and independent consultants. This is a massive capacity for nonformal surveillance if it can be properly marshaled, that has the potential to provide more surveillance samples than could be diagnosed. Therefore, diagnostic capacity must be maintained and improved as nonformal surveillance develops. Knowledge. Reaching this prospective surveillance force with timely and crop specific information is a significant challenge. While the aforementioned groups who typically participate in the CCA program have made great progress as information conduits, providing information to growers and other pest management players is a limiting factor in nonformal surveillance. Motivation. Growers are profit motivated and must attend to a complex set of details within a crop year if their bottom line is to be optimized. While some exotic pests may pose a threat to the current year’s yield, they may not be the bottom line threat that first comes to mind.

A summary of the challenges for each of the types of surveillance activities is found in Table 3.1. The remainder of the chapter will discuss how risk analysis, remote electronic or automated surveillance, increasing laboratory throughput, and diagnostic networks are integral to surveillance programs.

3.3 ESTIMATING RISK TO ORIENT SURVEILLANCE

The surveillance systems mentioned in the previous sections are highly dependent on human resources, which are dependent on the capacity, knowledge, and motivation of system players. Whether it is DHS and APHIS inspectors at points of entry, county extension agents monitoring sentinel plots, industry experts, university extension specialists conducting mobile surveillance, or farmers and their advisors walking into ﬁelds to take samples, a large number of trained people are required in the ﬁeld. For best results their efforts must be coordinated [21, 26] and guided by the best available information about risk. Given that the largest constraint to monitoring agricultural resources in the United States is the tremendous area that must be covered, and that human resources 48 DETECTION AND DIAGNOSIS OF PLANT PESTS

TABLE 3.1 The degree of challenge (high, medium, or low) associated with national work- force capacity, knowledge, and motivation related to plant pest detection strategies

Strategy Capacity Knowledge Motivation

Offshore and border High: extensive Medium: inter-and Medium: Agricultural monitoring for borders, coastline, intra governmental pest detection exclusion and import volume communications competes with other priorities Formal domestic High: large geographic Low: well established Low: local and monitoring area to cover formal agricultural state-based education system practitioners driven by local economy and client needs Nonformal domestic Medium: first detector High: large and Medium: profit impact monitoring volume can create diverse population of exotic pest(s) must overwhelming needing real-time be clear to first diagnostic surge information detectors

are limited, ideal monitoring systems must also rely on technology(ies) capable of antic- ipating threat and risk levels and automated detection processes. In this section we will discuss risk/threat evaluation techniques and remote sensing technologies that promote efficient and timely field-based surveillance by delivering biological and/or probability information prior to physical field scouting.

3.3.1 Threat analysis There are a range of approaches to threat analysis beginning with monitoring offshore pest movements [27], aerobiological modeling [28] and pathway analyses to determine possible modes of arrival [29–36], population dynamic models to assess threat of given plant pathogens as biological weapons [4], and disease progress models linked to climate models and geographic information systems (GIS) to predict probability of infection/establishment and direction of spread once introduced [3, 37].

3.3.2 Pathway Analyses Pathways of pest entry and spread are anticipated and assigned risk values by pathway analyses [38]. These analyses assign probabilities to natural entry, deliberate or unintentional introduction, establishment in the environment, rate of spread, and epidemic potential. Pathways can include commodity and seed trade, movement of plants and animals, conveyances that harbor hitchhiking biota; people who smuggle, travel with native plants or plant parts, trafﬁc in traditional medicinal plants, or intend to com- mit sabotage; and packing materials (particularly wood), potting media, even garbage. Target pests analyzed by APHIS are primarily regulatory threats as established by pest risk analyses. Knowing the most probable entry pathway(s) allows APHIS and DHS to focus interception efforts critical points. An indicator of deliberate introduction of a damaging pest would be if it were to occur outside of the most probable entry pathways [39]. ESTIMATING RISK TO ORIENT SURVEILLANCE 49

3.3.3 Models Plant pests and pathogens that are amenable to increase by scaled-up in vitro production, are stabile in storage, easy to transport and deliver are most likely to be developed as biological weapons. Whether the delivery results in establishment, dispersal, and damage, however, depends on the disease triangle: availability of susceptible host, climatic factors, and the inherent biology and physical characteristics of the organism [3, 4, 28]. The probability of any of these processes can be analyzed using mathematical models to assess risk of damage by a biological agent were it to be introduced intentionally. The most basic models used to assess potential impact of an introduced organism are “simple interest” (in the case of point source introductions of high levels of initial inoculum) or “compound interest” (where there would be cycles of inoculum build up over time) [3]. In the simple interest model, high risk is assigned to organisms that have highly effective inoculum (a high ratio of infection achieved over total inoculum propagules) over the broadest possible range of environments (which includes host genetics). For the compound interest model, high risk is assigned to those organisms that have a high rate of reproduction and infection over a broad range of possible environments. In this model, low amounts of inoculum could be introduced and would be hard to detect until an epidemic was fully blown, potentially causing impact in an area much larger than the original inoculation points. Modeling to assess risk of an invasive pest/pathogen (intentionally produced as a bioweapon or inadvertent), which establishes and damages the population growth rate (R), is the driving variable. Persistence of an organism over season can also be a function of R as overseason survival is a probability function of mortality of progeny. Thus high R pests and pathogens are generally assessed as higher risk of becoming endemic under appropriate climatic conditions, particularly if the dispersal is aerobiological. High R pests with a persistent overseasoning structure, that is a structure that ensures durability of the propagule, are the most likely to become established.

3.3.3.1 Aerobiological Models. Aerobiology is the study of the physical process of movement of organisms from one geographic location to another by floating, soaring, or flying through the atmosphere [28]. Pests and pathogens that disperse long distances readily are most likely to have wide-spread impact. Many plant pests and pathogens are physically designed for long-distance aerial transport, and these are subject to the dynamic but definable routes created by planetary airflows [28]. Aerobiological models are used to predict when and where a target organism might reach a location, and this helps establish the “normal” probability of appearance by natural processes. Many organisms that move long distances in the air, also move locally between the infected source and new locations. For the purpose of surveillance advisories, the probability of arrival of a given invasive organism to a new location can be forecast once host availability and climate variables are integrated with the knowledge of the biology of the organism. Likewise, an unknown source of origin can be estimated by a trace backwards of atmospheric pathways. The most widely used atmospheric transport model for aerobiological applications is the National Oceanic and Atmospheric Administration (NOAA)’s HYS- PLIT (hybrid single-particle Lagrangian integrated trajectory) model. Both APHIS and PIPE use this type of modeling [3, 24, 37].

3.3.3.2 Weather-based GIS and Disease/Pest Warning Models. Climate-based disease and pest forecast models are dependant upon accurate climatological data on a meaningful 50 DETECTION AND DIAGNOSIS OF PLANT PESTS

FIGURE 3.1 Mitigation efforts are most efficiently implemented when ground and remote surveillance inform aerobiological and HYSPLIT models to predict likely disease spread. geographic scale. Precision of climate data such as 24/7 temperature, dew point, and duration at or below dew point is a function of density of weather station data loggers and edaphic/topographic conditions of a region [40]. Relatively flat terrain such as the great plains and Midwestern region of the United States can expect reasonable precision of forecast with fewer data points over a larger scale. Conversely, complex topography can present highly diverse conditions on a fine scale. Thus, climate-based disease forecast in areas such as the Pacific northwest must take into account topography, altitude driven temperature and dew point variations, precipitation shadows, and so on. A denser network of weather data stations is required and interpolation between these cannot be generalized. Customized climate forecasts can be developed for specific production zones. Once climate model uncertainties and errors are understood, accurate disease severity forecasts can be generated based on biological characteristics of the organism or a proxy with similar attributes and known environmental parameters (Figure 3.1). The primary attribute that describes epidemiologic potential of an organism is its fitness or basic reproductive number (R), defined as a mean of new infections that result from an individual infection locus, or basically, the average number of progeny that an individual produces on a susceptible host under average (climatic) conditions [4, 41, 42]. Modelers understand how the R of each organism is affected by a range of climatic conditions. Some plant pathogens are highly dependent on climatic conditions during the process of infection of the host; so the probability of infection is a function of a threshold of effective inoculum and promoting climatic conditions. Other organism attributes that relate to R are length of latency or incubation, and short- to long-term progeny survival.

3.3.3.3 Population Dynamics Models:. Pest/pathogen population dynamics models can incorporate the final side of the triangle, the host. Host plant growth can be modeled based on factors such as phenology, degree days, and precipitation. The within- and between- season dynamic between host plant and pathogen can be simulated using reciprocal, coupled differential equation models [4, 44, 45]. The susceptible host is the essential ESTIMATING RISK TO ORIENT SURVEILLANCE 51 environment for the pathogen to develop, and successful development of the pathogen may have deleterious effects on its host. Strain composition and virulence diversity in populations tend to be well understood. Some pathogens are highly mutable and exist in populations polymorphic in virulence and aggressive on their host. These are managed by shifting host plant resistance to new varieties as resistance is defeated. In the presence of these types of pathogens, host resistance may break down gradually, over several years, or abruptly resulting in significant crop losses. Any sudden shift in virulence under these conditions could not be easily attributable; however, highly mutable pathogen populations are among some of the most destructive in the world. Other classes of pathogens have more stable population composition so that resistant varieties are more durable. Any sudden shift in virulence or aggressiveness of this type of pathogen would raise questions about how/why it happened, and often can be attributable to the specific introduction of a different strain.

3.3.3.4 Syndromic Analysis:. Intentional introduction of a plant pest or pathogen is likely to result in a pattern of disease outbreak that does not conform with the expected [3]. Pests and pathogens that are airborne and expected to fall out of air currents should distribute in a normal or random pattern across a predictable geographic swath. If the pattern of discovery is discrete and geographically distinct, this would fall outside of the expected and raise suspicion. Delivery of spores via a ground sprayer would result in a nonrandom disease presentation, such as in a line pattern along a fence row or road [3]. Other anomalies would be temporal and based on the likelihood of severity of outbreak at any given time. These anomalies are evaluated in real time on the basis of current diagnostic data, such as generated by NPDN labs (Figure 3.2), against the backdrop of historic data and/or climate-based forecast models of disease and pest outbreak. An outbreak that is too severe, too early in a growing season is a temporal anomaly [39]. Any departure from usual in disease or pest outbreak aggressiveness, severity, or incidence should trigger an investigation. Any change in intensity of mycotoxin production, particularly in the absence of a corresponding causal climatic factor, should be immediately investigated. It is paramount that analysts are trained, and automated analytical software developed to identify such anomalies, and that background data be archived in a format that is standardized for analysis.

3.3.4 Remote or Automated Surveillance Another way to supplement and focus monitoring and surveillance to narrow down the search parameters to manageable dimensions is to develop and deploy remote and/or automated sensing devices. This not only has the greatest immediate potential to supplement active surveillance capacity but also a long-term potential to supplement passive surveillance and exclusion. The research community is developing parameters required for detection, for example, specific nucleic acid, proteomic, or spectral signatures of specific organisms [6]. Tech- nologies for monitoring and notification about biological organisms in plant systems must be sensitive enough to identify the signature at very low concentrations, be associated with understanding of effective inoculum levels on available hosts (infectivity relative to environment), and be prompt enough to effect a successful control response. Ideally, such sensing devices could be designed to emit a signal to a remote monitor when a positive reaction or detection has occurred [45]. However, most sensors for environmental 52 DETECTION AND DIAGNOSIS OF PLANT PESTS

FIGURE 3.2 Syndromic analysis of data collected at state-based (triage) diadnostic laboratories. Triage labs conduct preliminary analysis. Regional and national labs perform confirmatory assays. biological monitoring still require at least some human manipulation such as substrate collection and transport for laboratory analysis. For example, fungal spores can be collected in outdoor aerial traps [46, 47], water traps [48], indoor suction traps [49] (Zhou). Krupa et al. have used the National Oceanic Atmospheric Deposition (NADP)/National Trends Network (NTN) water traps to monitor rain-deposited soybean rust in samples of rainwater [50, 51]. As few as two soybean rust spores can be detected with polymerase chain reaction (PCR) in rainwater samples collected in NOAA, National Atmospheric Deposition Program traps [51, 52]. Although spore capture has not yet been significantly correlated with crop disease, spores have been detected as far north as Canada. Nevertheless, atmospheric trapping of specific plant pathogens can only be a useful monitoring tool if there is no bottle neck in sample analysis. In the absence of in situ analysis systems, all remote trap samples must be collected and returned to the lab for processing. Therefore, the last aspect that we deal with in this paper is actually one of the most rate limiting aspects of all surveillance and monitoring programs, that of laboratory throughput capacity.

3.4 DIAGNOSTICS

3.4.1 Laboratory Throughput Whether by trapping, sentinel plot, mobile random survey, systematic survey, or by heightened awareness of a cadre of public and private first detectors, the consequence of regular monitoring can be an exponential increase in the numbers of samples requiring rapid diagnostic analysis [53]. When there is an emergency, and either a high-consequence biological agent or a new find of other high-consequence or quarantined pest has been detected, a surge of samples flowing into laboratories can be expected. Sampling needed to characterize and delimit an important outbreak, can result in a surge of samples and quickly overwhelm DIAGNOSTICS 53 the capacity of the laboratory to process the material opportunely. This is especially true when those who develop the sampling plans, for active surveillance either before or after first detection, do so without consideration of existing surge capacity. Surge management is dependent on human resources, physical space, reagent, and equipment availability within a laboratory [39, 54]. Table 3.2 shows a comparison of the three most common diagnostic methods for plant pathogen diagnosis. Traditional (incubation, microscopy, and taxonomic identification), molecular, and antibody-based diagnostics form the basis of the standard operating procedures (SOPs) that are developed for plant pathogen Select Agents and other high-consequence pathogens and pests. All plant diagnostic clinics are prepared for the traditional cultural diagnostic tests, and may use these as a first-order assay in the absence of a risk alert or other indication of unusual outbreak. PCR and immunoassay techniques are relatively quicker, more accurate, and preferred for high-consequence diagnostics. However, during a sample surge these techniques may be limited due to the level of training and testing (certifi- cation) required for personnel and inadequate supply of reagents [54]. Often, diagnostic confirmation requiring a high level of confidence will employ various technologies con- currently. An example of this was the SOP developed for diagnosis of the Sudden Oak Death pathogen, Phytophthora ramorum, which allowed for cultural, immunological, and nested PCR methods, depending on the level of capability and accreditation of the lab [55–57]. The challenges to large-scale screening of agricultural biological samples have been demonstrated both in simulated and actual surge events (refer to ICLN chapter). Indi- vidual laboratories, by themselves, can become overwhelmed in the case of a sudden influx of samples. Fulfilling the US biodefense strategy requires that laboratory response systems become capable of detecting biological agents from large numbers of samples quickly and cost effectively [54]. Alternative options exist, which range from investing in automated biological agent testing systems to increasing the throughput of national laboratories [53, 54, 58], to developing cascade protocols whereby diagnostic laboratories experiencing a surge can distribute samples over a network of laboratories significantly increasing the number of technical hours that can be brought to bear. Either of these options requires a national investment and would be effective regardless of the original cause of the surge (from intentional to natural).

3.4.2 US Laboratory Network System The NPDN [20] is a hub and spoke system of non-Federal laboratories distributed across the United States and territories, with varying capacity to effect determinant diagnoses (Figure 3.3). Every state and territory has at least one public diagnostic laboratory (Land Grant University plant clinic and/or State Department of Agriculture regulatory laboratory) which is a member of the NPDN. There are five regional hub laboratories that, in addition to processing samples submitted from within their state, coordinate diagnostics and information flow within each region. All of the laboratories are linked by a secured messaging system designed to alert diagnosticians of an increased risk of seeing a specific high-consequence biological agent. Protocols are in place for every state defining communications and information pathways and sampling flow operating procedures. By employing this network of existing public plant diagnostic laboratories, the United States has increased surge capacity for a moderate Federal investment. 54 DETECTION AND DIAGNOSIS OF PLANT PESTS

TABLE 3.2 Comparison of Diagnostic Methods

Variables Traditional Immunoassay PCR

Personnel needed 4 2 4 Analysis time 2–10 d 2 h 12 h Throughput/day >100 >200 Upfront costs Modest Moderate Sustained costs Low Low Confidence Agent may not grow on Are specific to genus Cross contamination media with high confidence, moderate, high can be very specific human error to viral strains Key characteristics Ultimate diagnostic tool Quickassaytoconfirm Flexible conditions for for confirmation presence/absence of assay development specific genera Pros Gold standard, positive Confirmation and High system identification of diagnosis inexpensive, flexibility, protocols organisms relatively simple; easily modified, monoclonal antibodies detects killed can be developed to organisms be general or specific, polyclonal antibodies can be developed quickly Cons Toxins not detected, Multiple steps increase Many opportunities very slow, may opportunity for for human error, require more human error flexibility increases extensive testing, the opportunity for agent may not grow applying wrong in artificial media, protocol unable to detect killed organisms Variability in access to standard pathogens/ clonal cell lines among manufacturers, more strain specificity with monoclonal antibodies, but there are no standard US cell lines.

Some NPDN laboratories are PCR trained and certiﬁed; others will be able to provide triage by examining and eliminating lower-consequence endemic pests and pathogens. However, when a triage laboratory receives a sample that it suspects to be an agent of national concern (whether homeland security or regulatory quarantine), the communications protocol for that state and region kicks into operation. One aspect of that protocol is to consult with the regional hub diagnostician and/or the speciﬁed Federal Reference Lab- oratory for the suspect agent. All presumptive positive samples of regulatory agents are DIAGNOSTICS 55

FIGURE 3.3 The “hub and spoke” structure of the NPDN allows: (i) regional coordination; (ii) surge overﬂow, and (iii) national sample triage.

National Identification Service labs – Confirmatory diagnosis

NPDN regional hub laboratories – Process samples, coordinate information with federal national labs.

NPDN state laboratories – Process samples, perform triage, return non-regulatory diagnostic results, refer suspect samples according to protocols.

Identification judgments by formal Identification judgments by first detectors. This includes surveillance programs conducted County extension agents, crop advisors, nursery through APHIS/PPQ, PIPE, etc. operators, crop producers who detect something new or unusual and deliver samples to the nearest laboratory

FIGURE 3.4 Diagnostic and triage hierarchy for US plant pest detection. forwarded to APHIS PPQ National Identification Services Laboratories for confirmatory diagnosis (Figure 3.4). All NPDN labs have digital diagnostics capability that allows real-time viewing of the same diagnostic features in two or more geographically separate laboratories via the Internet (Figure 3.5). This distance consultation provides first detectors and state diagnosticians local access to national diagnostic expertise in real time. Digital diagnostics have proven to reduce overall laboratory burden through triage, while improving diagnostic turn-around time on critical, presumptive high-consequence samples [23]. 56 DETECTION AND DIAGNOSIS OF PLANT PESTS

Nikon Eclipse DN100 ME-600 microscope digital network equipped for digital camera system photomicrography NPDN member laboratories

Monitor

NPDN hub laboratory

Remote unit USB Camera control mouse unit

Extension agents

First detectors

FIGURE 3.5 All NPDN member laboratories and properly equipped county extension ofﬁces can receive distance diagnostic support through digital image sharing on a state, regional, and national level.

3.5 CONCLUSIONS

The United States have over a billion acres of food, ﬁber, feed, and fuel production agriculture. Agriculture and forestry products combined are a cornerstone of the North American economy, providing social stability through food security and natural environments. Production agriculture is also being tapped to produce combustible fuels, exponentially increasing its overall value to producers and the American people. US agriculture could be considered a large, soft target to those who would want to strike at the economy, the social stability, or the sense of safety of the US citizenry. It would be technically easy to do this. Early detection and diagnosis of pests and diseases of plants is the best way to limit spread and impact, whether they arrive naturally or are introduced deliberately. This chapter has described the US plant pest detection and diagnostic systems. These systems are good, but vulnerability has increased due to globalization, terrorism, and the increasing value of agricultural products around the world. This increasing vulnerability will exacerbate current challenges. Therefore, it is in the country’s best interest to pursue research and infrastructural improvements to foster the security of that system.

3.5.1 Research

Monitoring and surveillance needs: • improved biologically based survey, sentinel, and sampling protocols • ﬁeld-based diagnostic kits • ongoing threat and pathway analyses REFERENCES 57

• epidemiology and aerobiology • syndromic analysis—the study of populations within communities and environments • remote/automated surveillance (microarrays and conductors). Diagnostic needs: • additional validated diagnostic assays • automation systems.

3.5.2 Infrastructure Planning and practice are essential to ensure an effective response to urgent public health threats [39]. This is true also for plant health threats. Integral to planning are education and policy. Needs are as follows:

• highly trained agricultural ofﬁcials at ports of entry whose mission is a Cabinet-level priority; • a new cadre of people educated in Agriculture Homeland Security that includes regulatory sciences, pathway and risk analysis, diagnostics, forensics; • ﬁeld workers, particularly growers, crop advisors, extension agents, who are knowledgeable about diseases that could be the result of use of biological agents, how to distinguish from the norm, and what to do when they encounter something suspi- cious. • quantitative biological scientists, epidemiologists, aerobiologists, and so on, for plant-based agriculture • alert and knowledgeable clinicians and laboratory diagnosticians, vital to disease surveillance efforts and recognition of new diseases and syndromes; and • automated laboratory high throughput equipment and operating procedures.

REFERENCES

1. Arnon, S. S., Schechter, R., Inglesby, T. V., Henderson, D. A., Bartlett, J. G., Ascher, M. S., Eitzen, E., Fine, A. D., Hauer, J., Layton, M., Lillibridge, S., Osterholm, M. T., O’Toole, T., Parker, G., Perl, T. M., Russel, P. K., Swerdlow, D. L., and Tonat, K. (2001). Botulinum toxin as a biological weapon: Medical and Public Health Management. JAMA 285(8), 1059–1070. 2. Dudley, J. P., and Woodford, M. H. (2002). Bioweapons, biodiversity, and ecocide: potential effects of biological weapons on biological diversity. Bioscience 52(7), 583–592. 3. Nutter, F. S., and Madden, L. V. (2005). Plant diseases as a possible consequence of biological attack. In M. S. Bronze, and R. A. Greenﬁeld, Eds. Biodefense: Principles and Pathogens, Horizon Bioscience, Norwich, Chapter 23. 4. Madden, L. V., and Vandenbosch, F. (2002). A population-dynamics approach to assess the threat of plant pathogens as biological weapons against annual crops. Bioscience 52(1), 65–74. 5. Madden, L. V., and Wheelis, M. (2003). The threat of plant pathogens as weapons against US crops. Annu. Rev. Phytopathol. 41, 155–176. 6. Fitch, P. J., Raber, E., and Imbro, D. R. (2003). Technology challenges in responding to biological or chemical attacks in the civilian sector. Science 302, 1350–1354. 58 DETECTION AND DIAGNOSIS OF PLANT PESTS

7. Gottwald, T., Graham, J., and Schubert, T. (2002). Citrus canker: the pathogen and its impact. Plant Health Prog. doi:10.1094/PHP-2002-0812-01-RV. 8. ARS (2004). Biology and Control of Emerald Ash Borer, Retrieved on May 30, 2008 from http://www.ars.usda.gov/research/projects/projects.htm?accn no=407426&showpars=true&fy= 2004. 9. Gilbert, T. (1996). Human Competence: Engineering Worthy Performance (Tribute edition), The International Society for Performance Improvement, Silver Spring, MD. 10. United States Department of Agriculture Animal and Plant Health Inspection Service (2007). Pest Detection. Retrieved March 30th, 2008 from http://www.aphis.usda.gov/plant health/ plant pest info/pest detection/index.shtml. 11. United States General Accounting Ofﬁce (2006). Homeland Security: Management and Coor- dination Problems Increase the Vulnerability of US Agriculture to Foreign Pests and Disease (GAO-06-044), Retrieved March 30th, 2008 from http://www.gao.gov/new.items/d06644.pdf. 12. Grode, J. (2005). American Seed Trade Association, Washington, DC. Retrieved March 30th, 2008 from http://www.amseed.com/ppts/Homeland Security.ppt#256,2,American%20 Seed%20Trade%20Association%20Washington, %20D.C. 13. United States Department of Agriculture Animal and Plant Health Inspection Service (2002). Questions and Answers About the Plant Protection Act, Retrieved March 30th from http://www.aphis.usda.gov/lpa/pubs/fsheet faq notice/faq phact.html. 14. Preslar, D. (2000). The Role of Disease Surveillance in the Watch for Agro-terrorism or Economic Sabotage, Retrieved March 30, 2008 from: http://www.fas.org/ahead/bwconcerns/ agroterror.htm. 15. National Plant Board (2002). Proceedings of 2002 National Plant Board 76th Annual Meet- ing, Duluth, MN Retrieved March 30th, 2008 from http://www.nationalplantboard.org/docs/ 2002 NPB Annual Meeting Proceedings Final.pdf. 16. Durbin, R. (2007). Senators Durban and Feinstein Introduce Bill to Transfer Responsibility for Agricultural Inspections from DHS back to USDA, Retrieved March, 30th, 2008 from http:// durbin.senate.gov/record.cfm?id=270685. 17. San Joaquin Farm Bureau Federation (2007). Ag wants Border Inspection Returned to USDA, Retrieved July 29th, 2007 from http://www.sjfb.org/thismonth/border.html. 18. United States Department of Agriculture Animal and Plant Health Inspection Service (2005). The Cooperative Agricultural Pest Survey: Detecting Plant Pests and Weeds Nationwide, Retrieved May 30, 2008 from http://www.aphis.usda.gov/publications/plant health/content/ printable version/pub phcapsdetecting.pdf. 19. Southern Region Integrated Pest Management Center (2008). IPM PIPE: About, Retrieved March 30th, 2008 from http://www.ipmpipe.org/about.cfm. 20. Stack, J., Cardwell, K. F., Hammerschmidt, R., Byrne, J., Loria, R., Snover-Clift, K., Baldwin, W., Wisler, G., Beck, H., Bostock, R., Thomas, C., and Luke, E. (2006). The National Plant Diagnostic Network. Plant Dis. 9, 128–136. 21. Southern Plant Board (2006). Resolution Number 8: Enhanced Funding of the Coopera- tive Agricultural Pest Survey Program Based on State-level Pest Introduction Risk Analysis, Retrieved March 30, 2008 from http://www.nationalplantboard.org/docs/resolution spb 2006 8 caps.pdf. 22. United States Department of Agriculture National Agricultural Statistics Service (2006). Wisconsin Pesticide Use, Retrieved March 30th, 2008 from http://www.nass.usda.gov/ Statistics by State/Wisconsin/Publications/Miscellaneous/pest use 06.pdf. 23. National Plant Diagnostic Network (2007). National Plant Diagnostic Network: a record of accomplishment, Retrieved May 30th, 2008 from https://www.npdn.org/library/viewdocument. pdf?ﬁletype=pdf&documentId=6431. REFERENCES 59

24. United States Department of Agriculture National Agriculture Statistics Service (2004). 2002 Census of Agriculture, Retrieved March 30th, 2008 from http://www.nass.usda.gov/census/ census02/volume1/us/CenV1US1.txt. 25. National Association of Independent Crop Consultants (2008). About the NAICC , Retrieved March 30th, 2008 from http://www.naicc.org/about naicc.cfm. 26. Isard S. A., Russo J. M., DeWolf E. D. (2006). The establishment of a national pest information platform for extension and education. Plant Health Prog. Retrieved May 30, 2008 from http://www.plantmanagementnetwork.org/php/elements/sum2.aspx?id=5508. 27. Balaam, R. J., United States Department of Agriculture Animal and Plant Health Inspec- tion Service Offshore Pest Information System (2004). Proceedings, XV USDA Interagency Research Forum on Gypsy Moth and Other Invasive Species. Retrieved May 30, 2008 from http://www.fs.fed.us/ne/newtown square/publications/technical reports/pdfs/2005/332% 20papers/balaam332.pdf. 28. Isard, S. A., Gage, S. H., Comtois, P., and Russo, J. M. (2005). Principles of the atmospheric pathway for invasive species applied to soybean rust. Bioscience 55, 851–861. 29. National Academy of Sciences (2002). Predicting Invasions of Nonindigenous Plants and Plant Pests, National Academy Press, Washington, DC, p. 194 (See chapter 2). 30. Andow, D. A. (2003). Pathway-based risk assessment of exotic species invasion. In Invasive Species, Vectors and Management Strategies,G.M.Ruiz,andJ.T.Carlton,Eds.IslandPress, Washington, DC, pp. 439–455. 31. Kiritani, K., and Yamamura, K. (2003). Exotic insects and their pathways for invasion. In Invasive Species, Vectors and Management Strategies,G.M.Ruiz,andJ.T.Carlton,Eds. Island Press, Washington, DC, pp. 44–67. 32. Mack, R. V. (2003). Global plant dispersal, naturalization, and invasion: pathways, modes and circumstances. In Invasive Species, Vectors and Management Strategies,G.M.Ruiz,and J. T. Carlton, Eds. Island Press, Washington, DC, pp. 3–30. 33. Auclairr, A. N., Fowler, G., Hennessey, M. K., Hogue, A. T., Keena, M., Lance, D. R., McDow- ell, R. M., Oryang, D. O., and Sawyer, A. J. (2005). Assessment of the risk of introduction of Anoplophora glavripennis (Coleoptera: Cerambycidae) in municipal solid waste from the quarantine area of New York City to landﬁlls outside of the quarantine area: a pathway analysis of the risk of spread and establishment. J. Econ. Entomol. 98(1), 47–60. 34. Work, T. T., McCullough, D. G., Cavey, J. F., and Komsa, R. (2005). Arrival rate of nonindigenous insect species into the United States through foreign trade. Biol. Invasions 7, 323–332. 35. Liebhold, A. M., and Work, T. T. McCullough, D. G., and Cavey, J. F. (2006). Airline baggage as a pathway for alien insect species invading the United States. Am. Entomol. 52(1), 48–54. 36. Worner, S. P., and Gevrey, M. (2006). Modelling global insect pest species assemblages to determine risk of invasion. J. Appl. Ecol. 43(5), 858–867. 37. Nutter, F. W. Jr, Rubsam, R. R., Taulor, S. E., Harri, J. A., and Esker, P. D. (2002). Geospatially-referenced disease and weather data to improve site speciﬁc forecasts for Stew- art’s will disease of corn in the U.S. corn Belt. Comput. Electron. Agric. 37, 7–14. 38. Hennessey, M. K. (2004). Quarantine pathway pest risk analysis at the APHIS Plant Epidemi- ology and Risk Analysis Laboratory. Weed Technol. 18, 1484–1485. 39. Rotz, L. D., and Hughes, J. M. (2004). Advances in detecting and responding to threats from bioterrorism and emerging infectious disease. Nat. Med. Suppl. 10(12), 130–136. 40. Coop, L. B. (2007). U. S. Degree-day Mapping Calculator, Version 3.0. Oregon State Uni- versity Integrated Plant Protection Center, Web Site Publication E.07-05-1: http://pnwpest.org/ cgi-bin/usmapmaker.pl. 60 DETECTION AND DIAGNOSIS OF PLANT PESTS

41. Zadoks, J. C. (1999). Reflections on space, time and diversity. Annu. Rev. Phytopathol. 7, 1–17. 42. Vanderplank, J. E. (1963). Plant Disease: Epidemics and Control, Academic press, San Diego. 43. Edelstein-Keshet, L. (1988). Mathematical Models in Biology, The Random House/Birkhauser mathematics Series, New York, NY. 44. Smith, J. M. (1971). Mathematical Ideas in Biology, Cambridge at the University Press, p. 152. 45. Babin, M., Cullen, J., Roesler, C. S., Donaghay, P. L., Douchette, G. J., Kahru, M., Lewis, M. R., Scholin, C. A., Sieracki, M. E., and Sosik, H. M. (2005). New approaches and technologies for observing harmful algal blooms. Oceanography 18, 210–227. 46. Schmale, D. G. III, Shah, D. A., and Bergstrom, G. C. (2005). Spatial patterns of viable spore deposition of Gibberella zeae in wheat fields. Phytopathology 95(5), 472–479. 47. Falacy, J. S., Grove, G. G., Mahaffee, W. F., Galloway, H., Glawe, D. A., Larsen, R. C., and Vandemark, G. J. (2007). Detection of Erysiphe necator in air samples using the polymerase chain reaction and species-specific primers. Phytopathology 97(10), 1290–1297. 48. Krupa, S., Bowersox, V., Claybrooke, R., Barnes, C., Szabo, L., Harlin, K., and Kurle, J. (2006). Introduction of soybean rust urediniospores into the Midwestern United States-a case study. Plant Dis. 9, 1254–1259. 49. Zhou, G., Whong, W.-Z., Ong, T., and Chen, B. (2000). Development of a fungus-specific PCR assay for detecting low-lever fungi in an indoor environment. Mol. Cell. Probes 14, 339–348. 50. Lamb, D., and Bowersox, V. (2000). The national atmospheric deposition program: an overview (2000). Atmos. Environ. 3(11), 1661–1663. 51. Barnes, C. W., Szabo, L. J., Isard, S. A., Ariatti, A., Tenuta, A. U., Hambleton, S., Tropiano, R., Bowersox, V. C., Claybrooke, R., and Lehmann, C., (2008). Patterns of Phakopsora pachyrhizi Spore Deposition Detected in North America Rain and Their Use to Calibrate IAMS Soybean Rust Forecasts in 2007 [abstract]. Phytopathology. 98, 518. 52. National Oceanic and Atmospheric Administration, Atmospheric Research Laboratory (2005). NOAA ARL HYSPLIT Model, Retrieved May 30, 2008 from www.arl.noaa.gov/ready/ hysplit4.html. 53. Byrne, K. M., Fruchey, I. R., Bailey, A. M., and Emanuel, P. A. (2003). Automated biological agent testing systems. Expert Rev. Mol. Diagn. 3(6), 759–768. 54. Emanuel, P. A., Fruchey, I. R., Bailey, A. M., Dang, J. L., Niyogi, K., Roos, J. W., Cullin, D., and Emanuel, D. C. (2005). Automated screening for biological weapons in homeland defense. Biosecur. Bioterror. 3(1), 39–50. 55. Osterbauer, N., Trippe A. (2005). Comparing Diagnostic Protocols for Phytophthora ramorum in Rhododendron, Plant Management Network, http://www.aphis.usda.gov/ppq/ispm/ pramorum/pdf files/pcrprotocol4.pdf. 56. USDA-APHIS. (2004). PCR Detection and DNA Isolation Methods for Use in the Phy- tophthora ramorum National Program, http://www.aphis.usda.gov/plant health/plant pest info/ pram/downloads/pdf files/cultureprotocol6-07.pdf. 57. USDA-APHIS. (2004). Guidelines for isolation by culture and morphological identification of Phytophthora ramorum, online: http://www.aphis.usda.gov/ppq/ispm/pramorum/pdf files/ pcrprotocol4.pdf. 58. Layne, S. P., and Beugelsdijk, T. J. (2003). High-throughput laboratories for homeland and national security. Biosecur. Bioterror. 2(1), 123–130. Voeller V05-c04.tex V1 - 12/05/2013 7:07pm Page 61

4 THE ROLE OF FOOD SAFETY IN FOOD SECURITY

Justin J. Kastner, Abbey L. Nutsch, and Curtis L. Kastner Kansas State University, Manhattan, Kansas

4.1 INTRODUCTION

Today’s agricultural security and food safety and security discussions are, admittedly, burdened by confusion of terminology. Therefore, this chapter necessarily begins by addressing the terms “security” and “defense” in the context of the agricultural and food industry. Some deﬁnitions that have been used by food regulators include the following:

• Food security. Activities associated with ensuring the adequacy of the food supply. • Food defense. Activities associated with protecting food from intentional contamination.

The term “food security” has been contested in recent years. In the post-9/11 era, new understandings of security have inﬂuenced the interpretation of both “homeland security” and “food security” [1, 2]. Indeed, the adoption, by such regulatory agencies as the US Food and Drug Administration, of the term “food defense” stems from confusion surrounding the term “food security” [3]. While some contend that “food security” ought to be strictly the domain of international-aid and economic-development policy communities, others have used the term “food security” to encompass international food defense issues as well as unintentional incidents that impact the adequacy of the food supply. Rather than letting semantics unduly complicate more important issues, leaders should support—or, at least, tolerate—the use of either “food security” or “food defense” in discussions devoted to how to protect the food supply and ensure food safety as well as food-supply sufﬁciency. Quite simply, knowledge of and practices regarding food safety are applicable regardless of one’s understanding of food security or food defense. The purpose of this chapter is not to debate the correct terminology but to show how food safety knowledge serves the broad objectives of food security/defense. Whether or not a food industry professional is concerned about bioterrorism, quality assurance, sanitation, physical site security,

61 Voeller V05-c04.tex V1 - 12/05/2013 7:07pm Page 62

62 THE ROLE OF FOOD SAFETY IN FOOD SECURITY border security, supply chain management, or international trade, he/she will ﬁnd that food safety capabilities and strategies are almost, if not always, applicable. The historical development of preventive, process-oriented food control systems notably includes the Hazard Analysis and Critical Control Point (HACCP) system; in the next section, the advent of HACCP and its application to food safety/defense is discussed. In order to paint a picture of the relevance of food safety principles, practices, and research to food security/defense, a situation- and commodity-speciﬁc example is offered in the next section. Finally, food safety education and its application to food security/defense is highlighted.

4.2 FOOD SAFETY PREVENTION, HACCP, AND FOOD SECURITY/DEFENSE

Food safety and control systems have developed in concert with better understanding of foodborne hazards [4]. Prior to the mid-nineteenth century, for example, a lack of epidemiological knowledge about microbiological pathogens left public health officials wondering what was the cause of diseases, including those brought on by the ingestion of particular foods. However, during the second half of the nineteenth century—and, indeed, up to the mid-twentieth century—particular foodborne illness cases were tied to particular agents (e.g. between 1850 and 1880, trichinosis and other diseases’ link to parasites was confirmed; between 1880 and 1950, specific bacterial diseases were tied to specific meat-borne pathogens). Between 1950 and 1985, food safety researchers continued to identify new microbiological (and chemical) hazards in food, and they became intrigued with the idea of intervening in food processing to reduce the likelihood (i.e. risk) of such hazards occurring [4]. By the mid-1980s, more deliberate process-oriented (as opposed to merely inspection-oriented) systems came into being [4]. One program—HACCP—had been developed previously to ensure food safety in the US Space Program; during the 1980s and 1990s, food companies and federal regulators began to insist on the seven principles of HACCP as a way to systematically control hazards (whether biological, chemical, or physical) in the food supply. The principles include the following:

1. Conduct a hazard analysis (identiﬁcation of biological, chemical, and physical hazards that may cause food to be unsafe). 2. Identify critical control points, that is, steps or procedures in which a hazard can be prevented, eliminated, or reduced. 3. Establish critical limits for each critical control point (e.g. speciﬁc temperature or processing parameters that ensure reduction of risk to an acceptable level). 4. Establish critical control point monitoring requirements. 5. Establish corrective actions in the event monitoring to indicate a violation of a critical limit. 6. Establish record keeping systems. 7. Establish validation procedures to demonstrate the HACCP system is in fact working.

HACCP is notably different from inspection-oriented food safety systems. HACCP’s introduction into the food industry has helped foster a prevention-oriented mindset Voeller V05-c04.tex V1 - 12/05/2013 7:07pm Page 63

CARVER + SHOCK AND THE EXAMPLE OF THE KANSAS MEAT INDUSTRY 63 amongst food professionals. Prevention-oriented food safety systems like HACCP are an invaluable asset in larger food security/defense efforts. Food safety systems can, in a sense, provide a kind of proverbial “downpayment” on ensuring that a food plant is culturally open to systematically considering, reducing, monitoring, and documenting risks of all kinds.

4.3 CARVER + SHOCK AND THE EXAMPLE OF THE KANSAS MEAT INDUSTRY

Examples are illustrative of wider concepts, and this section considers how food safety efforts can help address food-security vulnerability concerns. In 2006, Kansas State Uni- versity (K-State) conducted a vulnerability assessment of the meat processing industry in Kansas. The assessment revealed several instances in which food safety focused initiatives could be adapted to address food security/defense issues. Similarly, the assessment underscored how university-based food safety research could contribute to larger objectives of food security and food defense. This section explains the relevance of food safety efforts—including research efforts to broader activities designed to ensure a more secure food supply. CARVER + Shock is one of several tools and resources available through the US Food and Drug Administration [5]. CARVER is an acronym for six factors used to evaluate the attractiveness of a target for attack: Criticality (assesses the public health and economic impacts), Accessibility (the ability to physically access and egress from a target), Recuperability (ability of a system to recover from an attack), Vulnerability (ability to accomplish a successful attack), Effect (amount of direct loss from an attack as measures by loss of production), and Recognizability (ease of identifying a target). “Shock” assesses the combined health, economic, and psychological impacts of an attack within the food industry. Food safety systems (including HACCP) and food safety educational efforts (discussed at the end of this chapter) minimize food safety hazards and help provide an initial obsta- cle to achieving any significant consequences through intentional contamination. With regard to Criticality, the size of the meat processing industry in Kansas is consistently ranked among the top 3 in the nation. Indeed, the processing plants operated in Kansas by the world’s largest processors account for the bulk of that ranking; however, medium, small, and very small processors also contribute significantly to the high national ranking. Regardless of the relative contributions (i.e. whether production stems from large or small plants), Kansas is a major contributor to the totality of the United States and global meat industry. What happens in Kansas—whether in a large or small plant—impacts not only the state but also the entire domestic and international meat industry. Therefore, the safety of meat products processed in Kansas has far reaching implications that can impact both public health and economic stability. Regarding the safety and security of the meat supply, the US meat industry recognizes that all of its members could be “painted with the same brush” if meat safety were compromised. When considering meat processing along with allied livestock production, the total system is the number one revenue gen- erator in Kansas. The value of livestock, poultry, and their products in 2002 was $6.3 billion in Kansas alone [6]. The second element of CARVER + Shock—namely, Accessibility—also falls within the domain of traditional food safety efforts. Food safety officials are concerned about Voeller V05-c04.tex V1 - 12/05/2013 7:07pm Page 64

64 THE ROLE OF FOOD SAFETY IN FOOD SECURITY hazards and how they might be introduced into a food production system. The concept of food defense more purposefully addresses the issue of intentional actors who might introduce a hazard into the food supply. Terrorism, whether instigated by a transnational terrorist organization, a disgruntled employee, or a hired perpetrator, can result in similarly chaotic, catastrophic consequences for the food industry. Classically motivated terrorists prefer to destroy life as well as economic stability with an emphasis on taking as many lives as possible. Though the emphasis on protecting public health is a primary concern, an attack on the economy alone could, indeed, be devastating in and of itself. Because of the meat processing protocols being similar for beef, pork, lamb, and poultry, the following scenario was viewed as generically applicable to all species in the Kansas meat industry. Animals are normally held and rested for a period of time at the slaughter facility before being processed. A hazard could be introduced at that point and carried into processing. An externally administered hazard would be removed or at least diluted during dehiding, dehairing, pelting, or defeathering. Internalized hazards would be largely removed during evisceration. Therefore, to have the greatest impact, introduction of these later in processing was assessed as being more likely. Intact carcasses would be the next opportunity for cross contamination during slaughter or initial intentional contamination. However, this would primarily be limited to the surface of the carcass and subsequent food safety interventions (e.g. trimming, washing, and decontamination) would reduce or eliminate the hazard. It is the further processing of lean trim and particularly in comminuted systems (i.e. ground beef and wieners) that offers the greatest opportunity for introducing a hazard that would be uniformly incorporated into a large quantity of meat. Formulation ingredients (i.e. water and spices) would also widely distribute a hazard if contaminated. Further processing operations are readily accessible to workers and introduction of a hazard during mixing and blending would be relatively easy and effective. Smaller, isolated-yet-devastating episodes could be perpetrated at, for example, retail stores where grinding and mixing occurs [7]. With regard to the issue of Recuperability, food safety systems and food safety education place an emphasis on how to institute “corrective measures” in the event of a food safety violation. Traceability, mediated through attention-to-detail management and appropriate lot and date coding, would allow for product recall while the product is in storage or transit. All companies should have this capability to address food safety issues. However, if the hazard persisted undetected for a period of time, much of the product would be consumed prior to the recall. Even though it may not always be possible, consumers could use coding information to avoid consumption. Directions for disposal would also be required. Effective, appropriately crafted communications would be imperative to ensure an orderly and effective response. With regard to Vulnerability, those studying the Kansas meat industry concluded that a hazard introduction, while feasible, was likely to be less attractive than other segments of the food industry. (Indeed, if a widespread impact on public health was the terrorists’ goal, the beverage industry would be a better choice and terrorists have indicated their awareness of that.) Nevertheless, the impact of the livestock and meat industry on the economy could be dramatic. In fact, it was proposed that the following scenario would be feasible and have an impact on the economy. A hazard that both (a) has notable “shock value” and (b) is considered an adulterant (i.e. Escherichia coli O157:H7) in ground beef would, in the eyes of a terrorist, be a hazard of choice. The vulnerability assessors assumed that E. coli O157:H7 was evenly mixed by terrorists into a deﬁned lot Voeller V05-c04.tex V1 - 12/05/2013 7:07pm Page 65

CARVER + SHOCK AND THE EXAMPLE OF THE KANSAS MEAT INDUSTRY 65 of ground beef. Finished packages of that ground beef containing the lot identification would be taken by the terrorist(s). One week from that time a sample package would be submitted to the authorities. The product lot that is now on the retail shelves would then be verified as containing the hazard. Even though the hazard is diluted, and possibly of little or no public health consequence, consumer confidence in the US meat supply could be dramatically impacted. Indeed, the mere presence of E. coli O157:H7, regardless of the concentration, would initiate a recall of all the product in that lot and would impact the consumer demand for beef in general. The infective dose of E. coli O157:H7 is very low, and if the product was not cooked properly, it could cause foodborne illness. This scenario has been chronicled by the US Department of Homeland Security [8]. Additionally, terrorists might falsely claim that they had randomly contaminated more than the one lot and product type; under such a scenario, the economic impacts would be even further widespread. Large-scale meat recalls experienced by the US meat industry further illustrate the costs of such a scenario. In such scenarios, traditional food safety programs devoted to microbiological pathogens will be critical to minimizing vulnerability. Focusing on Effect, the vulnerability assessors discovered that if a meat product (processed by even a small meat processor in Kansas) could be verified as contaminated and is in the distribution chain, the effects would be dramatic. Even in the absence of a public health impact, the economic effects could be devastating. Here, food safety risk communication would be especially relevant. The vulnerability assessment did not address Recognizability in the sense of CARVER + Shock; officially, Recognizability relates to whether or not a perpetrator can identify the point for contamination. In addition, but in a different and sense of “recognizability,” the vulnerability assessors concluded that those interested in intentionally contaminating the food supply recognize the economic as well as potential public health impact. An in-line, real-time surveillance/detection device that could sense any level of any abnormal ingredient (i.e. the hazard) would be an ideal counter measure. Even though that does not currently exist, progress is being made in development of this capability. Prevention is preferred, but imperfectly effective in the real world. Therefore, a short-term strategy would be to train the work force to be aware of physical site security vulnerabilities, unusual behavior of co-workers, and security around processing areas of highest risk (e.g. mixing and blending operations). The assessors concluded that awareness, though not the ultimate answer for prevention, could help achieve that goal. To assist in this effort, a series of food safety and security modules on a variety of topics such as physical site security, threat recognition, response, and so on were made available in 2006 for Internet delivery from Kansas State University’s Division of Continuing Education. At this point, education to heighten awareness to recognize potential threats is an excellent first step. Dr. David Franz, an internationally recognized expert on bioterrorism, noted that education was the number one strategy based on effectiveness, including cost effectiveness [9]. With regard to the final element (Shock), the vulnerability assessors noted that an attack on the food supply could wreak economic, public health, and sociological havoc. The economy that the United States presently enjoys is implicitly reliant on a plentiful, affordable food supply. The magnitude of US discretionary spending is a result of an inexpensive food supply. Disruption of that food supply and its affordability would result in a chaotic situation that would rival any parallel loss of life [10]. For such sectors as the Voeller V05-c04.tex V1 - 12/05/2013 7:07pm Page 66

66 THE ROLE OF FOOD SAFETY IN FOOD SECURITY

Kansas meat industry, responsive food safety systems could, potentially, help minimize economic, public health, and mental health consequences.

4.3.1 Food Safety Research and Food Defense Local-area, university-based food safety research can help meet food defense/security concerns identified in the above CARVER + Shock vulnerability assessment. Shortly after the E. coli O157:H7 outbreak in the Northwest United States in 1993, K-State researchers set to work with an engineering firm and a major meat processor to help prevent such future safety outbreaks. In response to this food safety issue, steam pasteurization of beef carcasses was developed, validated, and widely employed in the meat processing industry for food safety purposes. Though directed at an incidental food safety event, the strategy of steam pasteurization of carcasses would also eliminate intentionally added hazards if they were susceptible to steam pasteurization. Assuming that an added hazard was eliminated by steam pasteurization, added hazards up to carcasses fabrication would be eliminated. Stated alternatively, a food safety solution developed by K-State researchers could help address food security/defense issues. To be sure, the opportunity exists for incidental as well as intentional contamination beyond other intervention strategies; in these cases, steam pasteurization of meat trimmings before grinding could prove valuable. By “hardening” (through the steam pasteurization of all beef carcasses and beef trim) a target, both food safety and food security/defense could be improved. This is but one example in which research to address food safety has come to address food security/defense concerns. Significantly, resources and practices serve a dual purpose and will make the food supply safer even if a terroristic event never occurs. K-State is but one of many universities across the United States where food safety programs are being used to address food security/defense issues. K-State is part of the Food Safety Consortium, a program that has been funded by the US Department of Agriculture since 1988 to address the safety of beef and beef products (at K-State), pork (at Iowa State University), and poultry (at the University of Arkansas). While the original Congressional mandate to the Consortium was to develop and validate methods and technologies to isolate, identify, and eliminate microbial and chemical hazards, those important areas have been enhanced through the addition of new research programs regarding the food safety challenges associated with food security/defense. Meanwhile, a joint K-State-New Mexico State University Frontier program for the historical studies of border security, food security, and trade policy (http://frontier.k-state.edu) seeks to inject interdisciplinary and social-science perspectives into traditional food safety and food security research. The National Center for Food Protection and Defense, based at the University of Minnesota, provides leadership in the overall food defense research and educational effort in the United States.

4.4 FOOD SAFETY EDUCATION IN FOOD SECURITY/DEFENSE

Traditional food safety related courses, such as food microbiology, food chemistry, epidemiology, toxicology, and so on are integral to food security/defense strategies and are Voeller V05-c04.tex V1 - 12/05/2013 7:07pm Page 67

FURTHER READING 67 part of most food science—related curricula. Food security/defense curricula can be easily augmented with these existing courses. Examples of this augmentation include a Food Safety and Defense Masters Certiﬁcate offered by K-State, the University of Nebraska at Lincoln, Iowa State University, and the University of Missouri. A similar initiative with Purdue University and the University of Indiana will lead to a graduate curriculum in Food Safety and Defense. These initial initiatives are being replicated and integrated with the educational efforts of the Department of Homeland Security Centers such as the University of Minnesota based National Center for Food Protection and Defense.

REFERENCES

1. Beresford, A. D. (2004). Homeland security as an American ideology: implications for U.S. policy and action. J. Homeland Secur. Emerg. Manag. 1(3), 301. 2. Kastner, J. and Ackleson, J. (2006). Chapter 6: global trade and food security: perspectives for the twenty-ﬁrst century. In Homeland Security: Protecting America’s Targets,J.J.F.Forest, Ed. Praeger Security International, Westport, CT and London, pp. 98–116. 3. FDA (2006). Food Defense Awareness: FDA Satellite Broadcast. 4. Koolmees, P. (2000). Chapter 4: Veterinary inspection and food hygiene in the twentieth century. In Food, Science, Policy and Regulation in the Twentieth Century,F.S.DavidandJ. Phillips, Eds. Routledge, New York, pp. 53–68. 5. U.S. Food and Drug Administration (2007). Food Defense and Terrorism, 10 December [cited 28 January 2008]. Available from: http://www.cfsan.fda.gov/∼dms/defterr.html 6. United States Department of Agriculture (2002). Kansas State and Country Data,Vol.1. Geographic Area Series Part 15 , National Agricultural Statistic Service. 7. Hui, Y. H., Hip, W.-K., Rogers, R. W., and Young, A., Eds. Meat Science and Applications, Marcel Dekker, Inc., New York, 2001. 8. U.S. Department of Homeland Security Risk Management Division Ofﬁce of Infrastructure Protection (2005). Characteristics and Common Vulnerabilities, Infrastructure Category, Beef Processing. 9. Franz, D. (2006). A multidisciplinary overview of food safety and security. Biological Security: An International Perspective (presentation of 18 May 2006, Kansas State University). 10. Jaax, J. (2006) A multidisciplinary overview of food safety and security. The Agricultural Bioterrorism Threat (presentation of 16 May 2006, Kansas State University).

FURTHER READING

Frazier, T. W. and Richardson, D. C., Eds. (1999). Food and Agricultural Security: Guarding Against Natural Threats and Terrorist Attacks Affecting Health, National Food Supplies, and Agricultural Economies, New York Academy of Sciences, New York. Voeller V05-c04.tex V1 - 12/05/2013 7:07pm Page 68 5 DECONTAMINATION AND DISPOSAL OF CONTAMINATED FOODS

M. Ellin Doyle, Seung Hak Lee, Craig H. Benson, and Michael W. Pariza University of Wisconsin, Madison, Wisconsin

5.1 INTRODUCTION

Widespread contamination of the food supply with a hazardous agent would be an effective way for a terrorist to induce panic in the general population and cause great economic losses. Food processing and distribution have become more centralized, such that contamination in one plant may result in multistate and even international outbreaks of illness as occurred in 2006–2007 with spinach contaminated with Escherichia coli O157:H7, peanut butter containing Salmonella, and pet food with melamine. Food companies have developed procedures for recalling foods containing undeclared allergens, foreign material, or pathogenic bacteria but may not be equipped to handle large volumes of foods containing a terrorist agent. Such an event could present a substantial waste disposal problem for landﬁlls and wastewater treatment plants (WWTPs) as well as for public health authorities. Representatives from disposal facilities, food companies, and government agencies participated in three meetings and voiced concern about a variety of issues that need to be addressed.

5.2 OVERVIEW

5.2.1 Agents Disposal options for contaminated foods will be determined to some extent by the nature and concentration of the agent and its expected fate in landﬁlls and wastewater treatment systems. Potential biological and chemical agents have been listed by Centers for Disease Control (http://www.bt.cdc.gov/agent/agentlistchem.asp, http://www.bt.cdc.gov/bioterrorism/) while toxic industrial chemicals that could pose a threat are listed by Occupational Safety and Health Administration (OSHA)

69 70 DECONTAMINATION AND DISPOSAL OF CONTAMINATED FOODS

(http://www.osha.gov/SLTC/emergencypreparedness/guides/chemical.html), and the most commonly encountered radionuclides are listed by Environmental Protection Agency (EPA) (Table 5.1). Some of these agents are inappropriate for poisoning food because their sensory characteristics would so alter the foods that they would not be consumed. Some microbes are not transmitted effectively through consumption of food and other agents are not soluble or stable in particular foods thereby precluding their use. There are also a number of other toxic chemicals and pathogens that terrorists may have access to, which are not on these lists. Accurate identiﬁcation of the agent(s) used to contaminate food and their concentrations is necessary for making appropriate disposal decisions.

5.2.2 Food A variety of foods could be contaminated by a terrorist and several factors, including the goal of the attack that will affect which food(s) might be chosen. If the goal is to cause widespread illness and death, then a highly infectious agent may be added to a commonly eaten, perishable food so that the agent will be widely dispersed and consumed before an alarm is raised. Massive economic losses and disruption of the food supply would be caused by contamination of animals in large rearing facilities or large areas of crop plants. Another goal may be to attack some American “icon” such as popular soft drinks, snacks, or fast foods. All of these possibilities should be considered in planning a response. Target foods must be reasonably accessible. Although many processing plants have instituted strict in-house security practices, places where liquids or other foods are mixed in large tanks or vats may provide an opportunity to deposit some hazardous agent into food. Imported foods and food ingredients may be more easily contaminated in their country of origin or during transport from overseas locations. An imported contaminated spice, for example, may then be added to many different foods. One potential weak link in many food supply chains occurs during transport from farms to food processors. A mathematical model simulating contamination of milk with botulinum toxin demonstrated that a single addition of toxin, occurring at a holding tank on a dairy farm, in a truck transporting milk to a processing facility, or in a raw milk silo at a processing plant, could poison up to 142,000 gallons of milk [2]. Quantity of food requiring disposal as well as its nature (solid, liquid, acidic and fatty) will affect disposal decisions. Relatively small amounts of food containing very hazardous agents may be incinerated but usually solid foods would be sent to a landﬁll and liquids to a wastewater treatment plant. Recent recalls of “naturally contaminated” foods have involved millions of pounds of ground beef and millions of cases of canned products (http://www.fsis.usda.gov/Fsis Recalls/index.asp) and a deliberate terrorist attack may contaminate even more food. If contamination is discovered while the food is still within control of the food processor, it may be contained as a point source, secured and stored while disposal decisions are made. But if contaminated food were already distributed to retail stores and sold to the public, then the hazardous agent could be widely dispersed throughout the country. Consumers would be likely to throw the food in the trash or pour it down the drain unless a convenient, efﬁcient collection system were organized to prevent indiscriminate disposal and further dissemination of the agent. TABLE 5.1 Radiological Agents: Potential Sources, Disposal Limits in Wastewater, and Partitioning in Incinerators

Estimated Partitioning (%)b

Chemical Concentration Combustion Radionuclide Formula Source Half-life Permitted (Bq/ml)a residues Fly ash

Americium 241Am Industrial/medical 432.7 yr 0.006 98 2 Cesium 137Cs Industrial/medical 30 yr 0.3 85 15 Cobalt 60Co Industrial/medical 5.3 yr 0.9 60 39 Iodine 129I 131I Industrial/medical 129I (15.7 × 106 yrs) 0.06 2 68 131I (8 d) 6.0 Iridium 192Ir Industrial/medical 73.83 d 3.0 Plutonium 238Pu Military/industrial 238Pu (87.7 yr) 0.006 98 2 239Pu 239Pu (24,100 yr) 0.006 240Pu 240Pu (6560 yr) 0.006 Polonium 210Po Military 138 d 0.012 Strontium 90Sr Industrial/medical 29.1 yr 0.15 95 5 Technetium 99Tc Industrial/medical 212,000 yr 18.0 40 50 Tritium 3H Military/industrial 12.3 yr – 0 0 Medical research Thorium 232Th Industrial 1.4 × 1010 yr 0.009 98 2 Uranium 238U Military/industrial 4.47 × 109 yr 0.09 98 2

See http://www.epa.gov/radiation/radionuclides/index.html. aMonthly average concentration limits for radionuclides disposed in sanitary sewers (10CFR20.2402). bPredicted partitioning of radionuclides in an incinerator [1]. 71 72 DECONTAMINATION AND DISPOSAL OF CONTAMINATED FOODS

5.2.3 Decontamination Neither municipal solid waste (MSW) landfills nor WWTPs would willingly accept foods with exotic terrorist agents if the fate of these agents during disposal were unknown. Effective decontamination or inactivation of agents may therefore be necessary before disposal. In a crisis situation, time constraints, cost, material limitations, and available diagnostic capabilities could limit decontamination options and it is not always clear what concentration of agent will be considered “clean enough” after decontamination [3]. Extensive information is available on destruction of many pathogens. However, some techniques are less effective when microbes are mixed with foods. EPA, the military, and some industries have had experience cleaning up chemical spills and deactivating chemical weapons and have developed decontamination procedures. Biological agents are generally susceptible to heat, irradiation, and disinfectants. Vegetative bacteria are readily killed by thermal treatments although the actual time/temperature requirements for different species vary somewhat depending on the concentration of fats, sugars, and salts [4]. Vegetative bacteria are also readily inactivated by ionizing radiation and by chlorine (residual chlorine of ∼1.1 mg/l) if the organic load is low [5]. Viruses, protozoan parasites, and biological toxins are somewhat more heat stable in foods but, with the exception of some toxins, they are inactivated by cooking and thermal processing of foods. Chlorine bleach solutions inactivate botulinum toxin, viruses, and vegetative bacteria on surfaces and in water but do not destroy protozoan parasites, mycotoxins, or ricin [6, 7]. Bacterial spores are much more resistant to environmental stress, including heat, desiccation, and sanitizers [8]. Bacillus anthracis spores can survive pasteurization and other thermal processing methods [9]. Autoclaving with steam and pressure ◦ (135 C, 217.2 kPa) for 40 min is generally effective in destroying all pathogens in medical waste. However, heat-resistant spores in building debris were completely destroyed only by two rounds of autoclaving [10]. Formaldehyde (final concentration of 5%) for 1–4 days can kill B. anthracis spores in liquid manure and sewage sludge (http://www.fas.org/nuke/intro/bw/whoemczdi986.htm). Ozone and ultraviolet (UV) light can also kill pathogens but high concentrations of organic matter in foods reduce their effectiveness significantly [11–13]. Chemical warfare agents can be inactivated by oxidizing agents such as hypochlorite and peroxides and by hydrolysis under alkaline or acidic conditions. Degradation products are usually less toxic than the parent compounds, but in some cases further treatment is necessary [14, 15]. Industries using toxic industrial chemicals have developed procedures to deal with accidental spills and releases of these chemicals. Radionuclides cannot be destroyed or inactivated. There are some methods, such as microfiltration and the use of magnetic particles to remove radionuclides from foods, which decrease the levels of contamination [16, 17] Following the Chernobyl accident, some of these procedures were used to remove radioisotopes in milk. Some processing methods significantly reduce radioactivity in edible parts of foods. For example, processing of milk into butter and cheese significantly reduces radiostrontium levels because this isotope partitions primarily into the aqueous fraction [18]. These methods would still leave radioactive materials that required disposal. FATE OF CONTAMINATED FOOD DURING DISPOSAL 73

5.3 FATE OF CONTAMINATED FOOD DURING DISPOSAL

5.3.1 Food The food matrix itself consists of organic compounds that would be easily degraded in landﬁlls, WWTPs, or incinerators. Since foods contain high moisture levels, their decomposition will produce relatively large amounts of leachate and may rapidly deplete oxygen and increase acidity in landﬁlls. Some liquids with a high organic load (such as milk) may disrupt microbial processes in activated sludge tanks and therefore may be added directly to an anaerobic digester in a wastewater treatment plant.

5.3.2 Biological Agents MSW landfills and WWTPs) regularly receive pathogenic bacteria, parasites, and viruses in human excreta, pet waste, disposable diapers, and spoiled foods. These are not a problem for well-run disposal facilities. But the presence of some bioterrorist agents in a large volume of food may present risks to workers and to public health if the agents are not inactivated or contained during the disposal process. Data are available on fate of some microorganisms under different disposal scenarios but few studies have investigated the fate of bioterrorist threat agents. Bacteria, viruses, and prions can attach to soil particles and potentially to landfill materials and may survive for extended periods in landfills [19–22]. Depending on the organism and landfill conditions, pathogens may also be carried in liquids that percolate down through the landfill material (leachate). Viruses appear to be the least hardy pathogens in landfills. No infective enteric viruses were cultured from fecally contaminated disposable diapers buried in landfills 2–10 years previously [23]. Hepatitis A virus and poliovirus survived more than 3 months in diapers ◦ ◦ and in landfill leachate at 5 C but were inactivated within a month at 40 C [24]. Viruses also persist in soil longer at cooler temperatures and when they adsorb to soil particles, ◦ but landfill temperatures are usually in the range of 40–50 C [19]. Enteric vegetative bacteria appear to survive longer than viruses in landfills with significant numbers of fecal organisms detected in 9–10-year-old landfill samples. Lysimeter studies with simulated refuse showed that survival of bacteria was related to rainfall, refuse content, temperature, and toxicity of leachate. Spores, however, are very resistant to environmental stress and are known to survive for centuries in soils [8]. Pathogens may attach to solid particles during wastewater treatment and settle into sludge. Size and surface properties of microbes as well as pH, ionic strength, and polyva- lent cation concentrations in the matrix determine attachment to particles [25]. WWTPs using only primary treatment remove about 12% of E. coli, 27% of Cryptosporidium oocysts, and negligible amounts of viruses indicating that many pathogens do not bind well to larger particles in settling tanks [26]. Some pathogens are destroyed during aerobic and anaerobic digestion of secondary/tertiary treatment processes but others survive typical wastewater treatment and are present in effluent and/or biosolids [27]. Chlorination and UV light can disinfect effluent before discharge but neither completely inactivate spores, oocysts, and all viruses [28]. Advanced processes, particularly chemical lime treatment combined with chlorination, can drastically reduce levels of bacteria, viruses, and parasitic oocysts in sludge [29]. 74 DECONTAMINATION AND DISPOSAL OF CONTAMINATED FOODS

During the outbreak of foot-and-mouth disease (FMD) in 2001 in the Netherlands, milk was illegally discharged into the sewer system. Since the virus might survive in WWTP and be present in efﬂuent, a quantitative risk assessment for FMD virus transmission to cattle drinking surface water was conducted. Based on available data and estimates for water and virus consumption by cows, it was concluded that discharge of contaminated milk into the sewer system could pose a high risk to cattle farms within 50 km of efﬂuent discharge to surface waters [30]. According to the EPA, more than 90% of medical waste that may contain infectious agents is currently incinerated because temperatures in properly run incinerators are hot enough to destroy all pathogens and toxins. Spore-forming bacteria can be killed within ◦ 30 min at 150 C while other bacteria and viruses are inactivated at temperatures of ◦ <100 C. However, an assessment by the EPA of the effectiveness of a medical waste incinerator in destroying heat-resistant bacterial spores demonstrated that some spores may survive in porous or moist materials where microenvironments do not reach the target temperature. In most cases, incineration caused >6 log kill of added spores but in some cases only about a 3 log kill occurred [31].

5.3.3 Chemical Agents Most chemical warfare agents and toxic industrial chemicals would be considered hazardous and food contaminated with them should be sent to hazardous waste landfills or incinerated. However, only about 21 commercial hazardous waste landfills and about 28 hazardous waste incinerators (that normally accept off-site generated wastes) are currently operating in the United States. Disposal of contaminated food in these hazardous waste facilities may not be a practical option due to distance from the contamination incident, volume of the food, and the need for rapid disposal. Physical and chemical properties of chemical agents deposited in MSW landfills or WWTP are important in determining whether they are volatilized, remain in aqueous or solid phases, or are inactivated by abiotic or biological reactions. Some studies on the behavior of chemical agents have been conducted in aqueous and soil systems but very limited information is available on the phase and fate of chemical agents in landfills and WWTP [32]. Although microbes possess enzymes that can degrade some chemical agents, it is not known whether they would significantly detoxify these agents under natural conditions in a landfill or WWTP [33]. Phase distribution and fate of some potential chemical terrorist agents in a landfill was modeled using Model for Organic Chemicals in Landfills (MOCLA)) [34]. Input parameters for the model include physical and chemical properties of the agents and landfill conditions derived from field data. The expected behavior of other chemical agents in landfills was estimated using MOCLA with similar landfill conditions and physicochem- ical properties obtained from the literature or estimated by EPI Suite v. 3.20 (Table 5.2). According to MOCLA, over 90% of most chemical agents, with the exception of sodium fluoroacetate, the cyanides, Amiton (VG), and VM would be distributed to the solid phase. Most of the remainder would be present in aqueous phase (leachate) with a negligible fraction in the gaseous phase. Even though a large fraction of most agents would be sorbed to the solid phase and therefore presumably be immobile, this does not guarantee that landfill disposal of these agents is safe. It merely represents the phase distribution of the chemicals remaining in a landfill. To address the total risk, the fate of the chemical agents, including transport and degradation, must also be considered. TABLE 5.2 Fate of Potential Chemical Terrorist Agents in Landfills and WWTP (estimated by MOCLA and STPWIN)

Landﬁll Phase Remaining in Distribution (%)* Landﬁll (%)* Estimated Fate in WWTP (%)

Bio Adsorption Present in Chemical Agent In water In solids 0.5 yr 5 yr degradation to sludge Volatilized efﬂuent

TICs Carbon disulﬁde 0.5 98.8 90.1 35.7 32.7 0.58 57.4 9.36 Furan 1.3 98.8 90.1 35.2 75.89 0.35 17.59 6.17 Blood Agents Potassium (sodium) cyanide 66.4 33.6 98.9 89.6 91.72 0.33 0 7.94 Sodium ﬂuoro-acetate (1080) 98.4 1.6 59.5 0.4 74.43 0.62 0.02 24.93 Incapacitating Agent Fentanyl 0.01 99.99 93.9 53.1 21.02 28.56 0 50.41 Anticoagulant Wafarin 0.1 99.9 93.3 50.1 48.12 2.56 0 49.32 Metals Thallium and its salts 7.6 92.4 99.4 94.3 Blister Agents Sulfur mustard (H/HD) 0.2 99.8 0 0 21.68 2.31 1.44 74.57 Lewisite (L) 0.2 99.8 0 0 20.96 2.57 7.51 68.96 Organo-phosphate Nerve Agents Tabun (GA) 5.9 94.1 0 0 74.46 0.62 0 24.92 Sarin(GB) 6.8 93.2 0.5 0 74.34 0.62 0.17 24.87 Soman (GD) 0.6 99.4 72.9 4.2 20.83 1.65 0.21 77.30 GE 3.5 96.5 7.3 0 74.32 0.63 0.23 24.82 Cyclosarin (GF) 0.8 99.2 52.9 0.2 74.55 0.72 0.18 24.56 Amiton (VG) 16.5 83.5 97.0 73.6 20.81 1.62 0 87.56 VM 30.1 69.9 93.6 51.7 20.63 1.50 0 87.87 VX 9.5 90.5 66.4 1.7 21.22 1.88 0 76.90

*Some data from Reference [34]. 75 76 DECONTAMINATION AND DISPOSAL OF CONTAMINATED FOODS

MOCLA calculates the fate of chemicals in a landfill, including advective transport of volatilized agents into the landfill gas collection system and through the soil cover, diffusive transport of volatilized agents through the soil cover, diffusive transport of soluble and volatilized agents through the composite liner, transformation of aqueous agents by biotic and/or abiotic mechanisms, and transport of soluble agents into the leachate collection system. A general half life of these compounds, obtained from EPI suite, was used in these simulations. This parameter includes oxidation/reduction, hydrolysis, and other reactions. Carbon disulfide and furan are predicted to move through landfills by gas-phase advection and according to Table 5.2, nearly 60 % of the initial amount would disappear within 5 years. At the end of landfill operations and 30 years of postclosure care, little of these chemicals would remain. However, attention should be directed to the gas monitoring/collection wells. MOCLA predicts that the cyanide agents will move through liquid-phase advection. Their disappearance rates are relatively slow due to their longer hydrolysis half life and lower Henry’s law constant. Nearly 90% would remain after 5 years with about 50% still present after 30 years. Leachate should be monitored for these compounds and properly treated. For most of the other chemicals, MOCLA indicates that hydrolysis will be the primary fate. Blister agents and most of the nerve agents degrade so rapidly that more than 95% of these chemicals would disappear from a landfill within 5 years. However, information is also needed on the fate of hydrolysis products. Previous studies have reported that some environmentally persistent hydrolysates, lewisite oxide and EA 2192 from VX, are highly toxic to mammals [35]. Thallium and other heavy metals are likely to persist for a long time in landfills and are known to concentrate in biosolids of WWTP. Fugacity-based analytical models can be used to predict the partitioning, transport, and transformation processes of organic chemicals in an activated-sludge wastewater treatment plant. EPA developed the Sewage Treatment Plant Fugacity Model (STPWIN) in EPI Suite v.3.20 to predict the multiphase partitioning and fate of chemicals in a sewage treatment plant (http://www.epa.gov/opptintr/exposure/pubs/episuite.htm). STPWIN was used to estimate the fate of chemical terrorist agents (Table 5.2). Most of the carbon disulfide and some of the furan would be volatilized while much of the fentanyl would be adsorbed on sludge particles. Microbes would degrade substantial amounts of the blood agents, some nerve agents, and warfarin. However, the model indicated that less than 30% of the blister agents and some nerve agents (GD, VG, VM and VX) would be removed and a large fraction of these chemicals would pass through the treatment plant intact and be present in effluent. Effluent would then require additional treatment to remove the residual chemicals. The major drawback of the STPWIN model is that abiotic transformation is not included in calculations but many blister and nerve agents are known to hydrolyze in aqueous systems with half-lives on the order of minutes and days. Some hydrolysis products are also toxic. Biodegradation rate is also a key parameter in this model but little information on the half-life of most chemical agents is available. Half lives for these chemicals were estimated using BIOWIN and the EPA draft method but any inaccuracy strongly affects STPWIN estimates of removal efficiency. The model also assumes complete mixing in the tanks fails to recognize the existence of layers or blankets of settled sludge. FATE OF CONTAMINATED FOOD DURING DISPOSAL 77

Hazardous wastes can be burned in specially designed incinerators that operate at higher temperatures than medical and municipal waste incinerators. A typical inciner- ◦ ator consists of a gas powered rotary kiln maintaining a temperature >980 C and an ◦ afterburner with a temperature <1200 C. Gases from the afterburner pass through an air pollution control system to remove particulates and other pollutants [36]. Incineration has been used by the US Army as a proven and reliable method for destroying the stockpile of chemical nerve and blister agents [37]. The ash and slag produced during incineration are sent to a hazardous waste landﬁll and emissions from the stacks pass through a pollution abatement system. A model simulation of destruction of chemical agents showed that, based on the incinerability indexes, an incinerator should be able to destroy chemical warfare agents easily and efﬁciently [38].

5.3.4 Radiological Agents Concentration of radioactivity and the isotopes present will determine disposal options for contaminated food. For radioisotopes with short half-lives, contaminated food may be stored securely until radioactivity decayed to very low levels. WWTP and MSW landfills may be acceptable sites for disposal of foods containing very low levels of radioactivity. The Nuclear Regulatory Commission permits discharge of water soluble radionuclides into sanitary sewer systems within certain limits (Table 5.1). If food contained greater levels of radioactivity or certain more toxic isotopes, alternative methods of disposal would be needed. Low levels of radioactivity, from naturally occurring isotopes, industrial waste, and excreta of people undergoing medical procedures, normally pass through WWTP. Some isotopes (131I, 60Co, 241Am, 40K, 226Ra, 228Ra, 89Sr, and 201Tl) are known to settle out with sludge particles and others have been reported to pass through the plant and are present in the discharged effluent [39, 40]. High levels of radionuclides, including 60Co used in cancer treatment and 241Am in industrial sources, were detected in sewage sludge in the past. In some cases, diluted radioactive materials were discharged into the sanitary system and diluted by millions of liters of liquid from other sources but radioisotopes became concentrated in sludge necessitating expensive cleanup operations [41]. Low level solid radioactive waste in the United States is normally deposited in one of the three operational radioactive waste landfills (Utah, Washington, and South Carolina). Both the concentration of the radionuclides and half-lives of the isotopes determine whether wastes are permitted in these landfills (http://www.nrc.gov/reading-rm/ doc-collections/cfr/part061/part061-0055.html). Potential radiological impact of disposal of large quantities of very low level solid radioactive waste (not more than 4 MBq/metric ton of β/γ activity in 108 kg/yr) in municipal landfills has been assessed by the UK government [42]. The assessment considered potential exposure of workers at a landfill site, the impact if the leachate from a landfill were directly deposited in a river, and problems that might occur after closing the site. Results from the study indicated that disposal of 3H, 36Cl, 90Sr, 238U, 239Pu or 241Am in landfills at this rate would be safe but disposal of 14C, 60Co, 137Cs, 226Ra and 232Th may not be safe. Incineration prior to disposal may be useful in reducing the expected large volume of contaminated food. Incineration does not destroy radioactivity but eliminates organic material, and radioisotopes are then present in a much smaller volume of residual ash, fused slag, fly ash, and volatile compounds. Expected partitioning of different isotopes 78 DECONTAMINATION AND DISPOSAL OF CONTAMINATED FOODS has been modeled (Table 5.1). Radioactive and toxic materials can be removed from gas emissions by scrubber filters and will require disposal. The radioactive ash can be immobilized in cement and then transported to a disposal or storage site [43]. Treatment capacities of controlled air incinerators for low level radioactive waste range from 200 to 700 kg/h.

5.4 CRITICAL NEEDS

Landfill operators have long-term responsibility and liability for their sites and depend on the good will of their neighbors and government agencies. WWTPs operate continuously and, as part of the public health system, have a very limited ability to refuse or redi- rect liquids that they receive. Neither would willingly accept foods containing terrorist agents without sufficient information about survival and fate of the agents. Escape of the agent from a landfill through gases, leachate, or leakage into groundwater is a major concern (Figure 5.1) as are the costs of increased monitoring for exotic agents, liability issues, risks to workers, and post-closure care and land use. If an agent is not degraded in a WWTP, it will be present in the effluent discharged to surface waters and/or in the biosolids normally spread on agricultural land. An agent may disrupt the metabolic activities of the bacteria performing the important tasks of degrading organic compounds, producing methane, and removing nitrogen from wastewater. Vapors or aerosols from activated sludge tanks may present a risk to workers (Figure 5.2). Perceived risk for some agents may be greater than the real risk but such issues must be addressed. Despite scientific evidence, neither the waste disposal operators nor the public may trust that the agent has been completely inactivated. Public health and public relations will require long-term monitoring to demonstrate that risk is minimal.

Landfill Disposal Issues Gas

F Storage Municipal Landfill Closure G postclosure Monofill, Special cell care

Truck A C General landfill cell

Decontamination D

Leachate E B

Hazardous Radioactive Incineration Discharge On-site Wastewater waste waste from treatment treatment landfill landfill liner to plant groundwater

FIGURE 5.1 Landﬁll disposal issues RESEARCH DIRECTIONS 79

Wastewater Disposal Issues

Sewage Water Treatment Plant pipes Activated Supernatant Effluent to sludge rivers, lakes C A Tank Sludge truck Anaerobic Biosolids digester

Municipal landfill Decontamination Hazardous waste landfill BD Radioactive waste landfill Agricultural Incineration land

FIGURE 5.2 Wastewater disposal issues

Critical needs include the following:

• greater coordination and communication between federal and state agencies that manage and regulate food and those that manage the environment and security; • development of regulations, policies, and procedures to deﬁne how terrorist wastes are to be managed, where they can be disposed, how they will be monitored, and what liabilities associated with managing these wastes will be assumed by the government and by industry; • more information is needed regarding the fate of terrorist agents and their degradation products in conventional pollution control systems, such as landﬁlls and WWTPs, so that rational decisions can be made regarding the impacts of accepting contaminated foods; • planning by pollution control facilities to ensure worker safety and address the potential for inadvertent releases to ensure public safety and protection of the environment; • increased communication among all stakeholders including food companies, pollution control facilities, renderers, transportation industry, and diagnostic laboratories; • risk management strategies must be developed to communicate openly and honestly with the public.

5.5 RESEARCH DIRECTIONS

Research on the fate of various terrorist agents in pollution control systems should be investigated soon before the society and disposal industry are faced with a daunting 80 DECONTAMINATION AND DISPOSAL OF CONTAMINATED FOODS disposal problem. A sound technical basis is needed so that rational decisions can be made about the disposal of large volumes of food contaminated by terrorists. Future research should be directed to the following: • investigation of the fate of various agents in different food matrices under conditions similar to those encountered in landﬁlls and WWTP; • improved analytical methods for various agents in different food matrices; • feasible methods for decontaminating large volumes of contaminated food; • effective risk communication strategies.

REFERENCES

1. Burger, L. L. (1995). A Chemical Basis for the Partitioning of Radionuclides in Incinerator Operation, PNL-10364, UC-601, U.S. Department of Energy. 2. Wein, L. M., and Liu, Y. F. (2005). Analyzing a bioterror attack on the food supply: the case of botulinum toxin in milk. Proc. Natl. Acad. Sci. U.S.A. 102, 9984–9989. 3. Raber, E., Carlsen, T. M., Folks, K. J., Kirvel, R. D., Daniels, J. I., and Bogen, K. T. (2004). How clean is clean enough? Recent developments in response to threats posed by chemical and biological warfare agents. Int. J. Environ. Health Res. 14, 31–41. 4. Doyle, M. E., and Mazzotta, A. S. (2000). Review of studies on the thermal resistance of salmonellae. J. Food Prot. 63, 779–795. 5. Rose, L. J., Rice, E. W., Jensen, B., Murga, R., Peterson, A., Donlan, R. M., and Arduino, M. J. (2005). Chlorine inactivation of bacterial bioterrorism agents. Appl. Environ. Microbiol. 71, 566–568. 6. Henghold, W. B. (2004). Other biologic toxin bioweapons: ricin, staphylococcal enterotoxin B, and trichothecene mycotoxins. Dermatol. Clin. 22(3), 257–262. 7. Villar, R. G., Elliott, S. P., and Davenport, K. M. (2006). Botulism: the many faces of botulinum toxin and its potential for bioterrorism. Infect. Dis. Clin. North Am. 20, 313–327. 8. Nicholson, W. L., Munakata, N., Horneck, G., Melosh, H. J., and Setlow, P. (2000). Resistance of Bacillus endospores to extreme terrestrial and extra-terrestrial environments. Microbiol. Mol. Biol. Rev. 64, 548–572. 9. Novak, J. S., Call, J., Tomasula, P., and Luchansky, J. B. (2005). An assessment of pasteurization treatment of water, media, and milk with respect to Bacillus spores. J. Food Prot. 68, 751–757. 10. Lemieux, P., Sieber, R., Osborne, A., and Woodard, A. (2006). Destruction of spores on building decontamination residue in a commercial autoclave. Appl. Environ. Microbiol. 72, 7687–7693. 11. Blatchley, E. R., Meeusen, A., Aronson, A. I., and Brewster, L. (2005). Inactivation of Bacillus spores by ultraviolet or gamma radiation. J. Environ. Eng. 131, 1245–1252. 12. Lytle, C. D., and Sagripanti, J. L. (2005). Predicted inactivation of viruses of relevance to biodefense by solar radiation. J. Virol. 79, 14244–14252. 13. Mahapatra, A. K., Muthukumarappan, K., and Julson, J. L. (2005). Applications of ozone, bac- teriocins and irradiation in food processing: a review. Crit. Rev. Food Sci. Nutr. 45, 447–461. 14. Talmage, S. S., Watson, A. P., Hauschild, V., Munro, N. B., and King, J. (2007). Chemical warfare agent degradation and decontamination. Curr. Org. Chem. 11, 285– 298. REFERENCES 81

15. Wagner, G. W., and Yang, Y. C. (2002). Rapid nucleophilic/oxidative decontamination of chemical warfare agents. Ind. Eng. Chem. Res. 41, 1925–1928. 16. Gao, Y., Zhao, J., Zhang, G. H., Zhang, D., Cheng, W. W., Yuan, G. Q., Liu, X. J., Ma, B. Z., Zeng, J. H., and Gu, P. (2004). Treatment of the wastewater containing low-level Am-241 using flocculation-microfiltration process. Sep. Purif. Technol. 40, 183–189. 17. Navratil, J. D. (2001). Pre-analysis separation and concentration of actinides in groundwater using a magnetic filtration/sorption method I. background and concept. J. Radioanal. Nucl. Chem. 248, 571–574. 18. Patel, A. A., and Prasad, S. R. (1993). Decontamination of radioactive milk—a review. Int. J. Radiat. Biol. 63, 405–412. 19. Davies, C. M., Logan, M. R., Rothwell, V. J., Krogh, M., Ferguson, C. M., Charles, K., Deere, D. A., and Ashbolt, N. J. (2006). Soil inactivation of DNA viruses in septic seepage. J. Appl. Microbiol. 100, 365–374. 20. Ma, X., Benson, C. H., Mckenzie, D., Aiken, J. M., and Pedersen, J. A. (2007). Adsorption of pathogenic prion protein to quartz sand. Environ. Sci. Technol. 41, 2324–2330. 21. Morrow, J. B., Stratton, R., Yang, H. H., Smets, B. F., and Grasso, D. (2005). Macro- and nanoscale observations of adhesive behavior for several E. coli strains (O157:H7 and environmental isolates) on mineral surfaces. Environ. Sci. Technol. 39, 6395–6404. 22. Rzezutka, A., and Cook, N. (2004). Survival of human enteric viruses in the environment and food. FEMS Microbiol. Rev. 28, 441–453. 23. Huber, M. S., Gerba, C. P., Abbaszadegan, M., Robinson, J. A., and Bradford, S. M. (1994). Study of persistence of enteric viruses in landfilled disposable diapers. Environ. Sci. Technol. 28, 1767–1772. 24. Gray, M., Deleon, R., Tepper, B. E., and Sobsey, M. D. (1993). Survival of hepatitis-A virus (Hav), poliovirus-1 and F-specific coliphages in disposable diapers and landfill leachates. Water Sci. Technol. 27, 429–432. 25. Crockett, C. S. (2007). The role of wastewater treatment in protecting water supplies against emerging pathogens. Water Environ. Res. 79, 221–232. 26. Payment, P., Plante, R., and Cejka, P. (2001). Removal of indicator bacteria, human enteric viruses, Giardia cysts, and Cryptosporidium oocysts at a large wastewater primary treatment facility. Can. J. Microbiol. 47, 188–193. 27. Shannon, K. E., Lee, D. Y., Trevors, J. T., and Beaudett, L. A. (2007). Application of real time quantitative PCR for the detection of selected bacterial pathogens during municipal wastewater treatment. Sci. Total Environ. 382, 121–129. 28. Blatchley, E. R., Gong, W. L., Alleman, J. E., Rose, J. B., Huffman, D. E., Otaki, M., and Lisle, J. T. (2007). Effects of wastewater disinfection on waterborne bacteria and viruses. Water Environ. Res. 79, 81–92. 29. Rose, J. B., Huffman, D. E., Riley, K., Farrah, S. R., Lukasik, J. O., and Hamann, C. L. (2001). Reduction of enteric microorganisms at the upper Occoquan Sewage Authority water reclamation plant. Water Environ. Res. 73, 711–720. 30. Schijven, J., Rijs, G. B. J., and Husman, A. (2005). Quantitative risk assessment of FMD virus transmission via water. Risk. Anal. 25, 13–21. 31. Wood, J. (2006). Thermal destruction of Bacillus anthracis surrogates in a pilot scale incinerator. Air and Waste Management Association, 99th Annual Conference, New Orleans, paper #328, 12. 32. Kingery, A. F., and Allen, H. E. (1995). The environmental fate of organo-phosphorus nerve agents: a review. Toxicol. Environ. Chem. 47, 155–184. 33. Singh, B. K., and Walker, A. (2006). Microbial degradation of organophosphorus compounds. FEMS Microbiol. Rev. 30, 428–471. 82 DECONTAMINATION AND DISPOSAL OF CONTAMINATED FOODS

34. Bartelt-Hunt, S. L., Barlaz, M. A., Knappe, D. R., and Kjeldsen, P. (2006). Fate of chemical warfare agents and toxic industrial chemicals in landfills. Environ. Sci. Technol. 40, 4219–4225. 35. Munro, N. B., Talmage, S. S., Griffin, G. D., Waters, L. C., Watson, A. P., King, J. F., and Hauschild, V. (1999). The sources, fate, and toxicity of chemical warfare agent degradation products. Environ. Health Perspect. 107, 933–974. 36. Santoleri, J. J., Reynolds, J., and Theodore, L. (2000). Introduction to Hazardous Waste Incin- eration. Wiley-Interscience, New York. 37. National Research Council (U.S.) (2007). Committee on review of chemical agent secondary waste disposal and regulatory requirements. Review of Chemical Agent Secondary Waste Dis- posal and Regulatory Requirements. National Research Council, Washington, DC. 38. Denison, M. K., Sadler, B. A., Montgomery, C. J., Sarofim, A. F., Bockelie, M. J. (2004). Computational modeling of a chemical liquid incinerator chamber. IT3’04 Conference, May 10–14, 2004 , Phoenix, Arizona. 39. Bastian, R. K., Bachmaier, J. T., Schmidt, D. W., Salomon, S. N., Jones, A., Chiu, W. A., Setlow, L. W., Wolbarst, A. B., Yu, C., Goodman, J., and Lenhart, T. (2005). Radioactive materials in biosolids: national survey, dose modeling, and publicly owned treatment works (POTW) guidance. J. Environ. Qual. 34, 64–74. 40. Martin, J. E., and Fenner, F. D. (1997). Radioactivity in municipal sewage and sludge. Public Health Rep. 112, 308–316. 41. General Accounting Office (US). (1994). Radionuclides at Sewerage Treatment Plants, GAO/RCEL-94-133. GAO, Gaithersburg, MD. 42. Chen, Q. Q., Kowe, R., Mobbs, S. F., and Jones, K. A. (2007). Radiological Assessment of Disposal of Large Quantities of Very Low Level Waste in Landfill Sites, http://www.hpa.org.uk/ radiation/publications/hpa rpd reports/2007/hpa rpd 020.pdf. 43. International Atomic Energy Agency (2004). Predisposal Management of Organic Radioactive Waste , (IAEA-TECHDOC-427), IAEA, Vienna, Austria.

FURTHER READING

Clark, B., Henry, G. L. H., and Mackay, D. (1995). Fugacity analysis and model of organic chemical fate in a sewage treatment plant. Environ. Sci. Technol. 29, 1488–1494. Food Safety Department, World Health Organization (2002). Terrorist Threats to Food: Guid- ance for Establishing and Strengthening Prevention and Response Systems, http://www.who.int/ foodsafety/publications/general/en/terrorist.pdf. International Atomic Energy Agency (2006). Application of Thermal Technologies for Processing of Radioactive Waste, (IAEA-TECHDOC-1527), IAEA, Vienna, Austria. Lemieux, P., Thorneloe, S., Nickel, K., and Rodgers, M. (2007). A Decision Support Tool (Dst) for Disposal of Residual Materials Resulting from National Emergencies, http://www.epa.gov/ nhsrc/pubs/paperDSTDispResidual101007.pdf. Lund, B. M., Baird-Parker, T. C., and Gould, G. W. (2000). The Microbiological Safety and Quality of Food. Aspen Publishers Inc., Gaithersburg, MD. Palmisano, A. C., and Barlaz, M. A. eds (1996). Microbiology of Solid Waste. CRC Press, New York. 6 PULSENET: A PROGRAM TO DETECT AND TRACK FOOD CONTAMINATION EVENTS

Kara L. F. Cooper, Duncan R. MacCannell, and Efrain M. Ribot Centers for Disease Control and Prevention, Atlanta, Georgia

6.1 INTRODUCTION

An estimated 76 million cases of foodborne illness occur each year in the United States [1], with many more that likely go unreported. The striking incidence of foodborne disease highlights the importance of effective surveillance for the rapid detection of outbreak clusters, and the role that these programs have in maintaining the safety of our food supply [2–5]. In recent years, a number of well-publicized outbreaks have been associated with widely distributed food products. Each outbreak underscores the potential vulnerability of our food supplies to microbial contamination, the threat to public health, and the potential impact to national and international security. Contamination can occur at any stage of food production, from the originating farm or producer, through processing, packaging, shipping and storage, all the way to the consumer’s table. Although changes in regulation, education, and technology over the course of the past century have resulted in an overall improvement in the safety of the foods we eat [2, 4], changing dietary habits and demand for increased variety and seasonality of many food products presents an important challenge to food safety, particularly when many foods are imported from abroad and/or consumed in an uncooked format. Today, many commercially prepared foods are produced or processed at centralized facilities and are widely distributed prior to consumption. Both of these factors facilitate the emergence of disseminated outbreaks, and foodborne contamination may impact multiple counties, states, or countries [6]. Furthermore, the complexity of these production and distribution networks can delay the recognition of foodborne outbreaks or hinder the identiﬁcation and recall of contaminated products [2–4, 7]. The rapid detection of disease clusters is the critical function of PulseNet, and their association with outbreaks of foodborne disease combines traditional epidemiology with real-time, laboratory-based surveillance systems. In order to be effective, it also requires sufﬁcient coverage, and the active participation of laboratorians and epidemiologists at both the state and federal levels [3,8,9].

83 84 PULSENET: A PROGRAM TO DETECT AND TRACK FOOD CONTAMINATION EVENTS

TABLE 6.1 Examples of International Outbreaks of Bacterial Foodborne Outbreaks Rec- ognized through PulseNet, 1995–2006

Year Organism Cases Countries Involved Implicated Vehicle Reference

1995 S. Stanley 250+ Finland, USA Alfalfa sprouts Mahon [51] 1998 Shigella sonnei 172 Canada, USA Parsley MMWR [31] 1999 S. Muenchen 207 Canada, USA Unpasteurized Orange MMWR Juice [31] 2000–2001 S. Enteritidis 168 Canada, USA Raw Almonds Chan [28] 2000–2002 S. Poona Canada, Mexico, Cantaloupe MMWR USA [29] 2001 S. Oranienburg 500+ Austria, Belgium, Chocolate Werber [27] Denmark Finland, Germany, Croatia Netherlands, Czech Repulic 2004 S. Enteritidis 29 Canada, USA Raw Almonds MMWR [30] 2004 E. coli 3 Japan, USA Ground Beef MMWR O157:H7 [52] 2006 E. coli 200+ Canada, USA Fresh spinach MMWR O157:H7 [25]

PulseNet USA was established in 1996 to provide a unified infrastructure for the molecular surveillance of foodborne bacterial infections, and has significantly increased our ability to detect and act upon clusters of foodborne illness by standardizing and coordinating bacterial subtyping in public health laboratories nationwide [10, 11]. Participating laboratories compare DNA fingerprints from foodborne isolates across the country in near real time, and can compare these fingerprints to those from a suspected vehicle or source during the course of an investigation. This secondary capability is particularly useful during traceback investigations, and is critical to both attribution and prevention/control efforts [9, 11, 12]. Hundreds of outbreak clusters are identified by PulseNet laboratories each year, prompting dozens of outbreak investigations. From the perspective of biosecurity and preparedness, many of the lessons learned from the investigation and follow up of these cases are equally applicable to scenarios involving intentional contamination. The PulseNet community is committed to improving the capacity and technology needed to respond rapidly to emerging threats and limit the scope of foodborne infections, both within the United States and abroad.

6.2 SCIENTIFIC OVERVIEW

Molecular epidemiology combines the principles of traditional epidemiologic investigation with tools derived from molecular biology to better understand and define the distribution of diseases or markers of interest within a given population [7]. As technology PULSENET USA 85 continues to improve, the molecular strain typing of bacteria has evolved significantly, particularly with the introduction and refinement of modern genomic and proteomic approaches [13]. Today, epidemiologic studies of foodborne bacteria rely predominantly upon genotypic methods for characterization and strain type assignment. Although the sensitivity, specificity, and appropriateness of these techniques varies according to both organism and context, pulsed-field gel electrophoresis (PFGE), which involves macrorestriction and fragment analysis of the entire genome, remains the predominant means or “gold standard” for subtyping many bacteria [13, 14]. The application of sophisticated molecular subtyping methods, coupled with informatic platforms that enable rapid comparison and communication of subtyping data, has greatly increased the sensitivity of surveillance systems to detect clusters of foodborne illness that are widely dispersed in space or time [3, 11]. However, detection of a cluster is only the first step, and a rapid but thorough investigation is critical to identify the source of infection and mediate an effective response before additional and preventable cases occur. Molecular subtyping data is often central to this process, helping to separate outbreak-related cases from sporadic infections that would otherwise confound the investigation. Once a contaminated product or vehicle has been identified, molecular data may also be used to provide definitive microbiological confirmation of the source of the infection [2, 4, 7, 9, 12]. Equally, these data may be used in a detailed investigation of affected food production and distribution chains, which may reveal how contamination occurred, and facilitate the development and implementation of measures to prevent similar events in future.

6.3 PULSENET USA

PulseNet USA was established in 1996 in collaboration with the Association of Public Health Laboratories, and has grown from an initial group of five interested public health laboratories to over 70 participants, including county, municipal and state public health departments, federal food regulatory agencies, agricultural, and veterinary laboratories [12]. Standardized PFGE protocols are at the core of PulseNet USA, allowing participating laboratories to rapidly generate PFGE patterns or “fingerprints” that are consistent between laboratories [13, 15, 16]. Once an isolate has been run, the PFGE pattern is compared to other local isolates and uploaded to a centralized and secure database system at the Centers for Disease Control (CDC) in Atlanta. These pattern data are actively curated, with ongoing real-time surveillance for unusual pattern frequencies or trends that may indicate the emergence of an outbreak. The participation of laboratories throughout the country enables the exquisitely sensitive and rapid identification of foodborne outbreak clusters, particularly those that are disseminated over a broad geographic area or protracted period of time. An important key to the success of PulseNet USA has been the decentralization of subtyping activities to state and local public health laboratories. This decentralized approach has greatly extended the testing capacity of the network, with rapid subtyping occurring on a local level, and without the logistical problems of a centralized testing facility [11, 12]. Decentralization presents its own challenges, however: in particular, the allocation of infrastructure (i.e. equipment, supplies, and trained personnel) and funding, the development and maintenance of highly standardized subtyping and analysis protocols 86 PULSENET: A PROGRAM TO DETECT AND TRACK FOOD CONTAMINATION EVENTS for pathogens of interest, and a means of effective, rapid, and secure communication for data and alerts between network members. PulseNet relies on the ability to compare data generated by different laboratories in near-real time, and it is therefore necessary to ensure that protocols are standardized, and that data may be rapidly exchanged and compared. Within the PulseNet community, PFGE gel images are analyzed using highly customized image analysis software (BioNumerics, Applied Maths, Sint-Martens-Latem, Belgium), that operate using a basic client–server architecture. Through this software, PFGE patterns and related epidemiological information can be compared against both the local (client) database and against all patterns that have been submitted to the National database (server) at the CDC. Currently, PulseNet has standardized PFGE protocols and networked databases in place for Escherichia coli O157, Salmonella, Shigella spp., Campylobacter, Listeria monocytogenes, Vibrio cholerae, Yersinia pestis, and Vibrio parahaemolyticus [16–21]. To provide rapid and secure messaging between PulseNet laboratories, a web-based forum has also been implemented to provide rapid alerts to laboratorians and epidemiologists throughout the country, and to provide timely information to the PulseNet community on patterns, tentative clusters, and important administrative or methodological updates. Together, the combination of highly standardized molecular protocols, database-driven curated analysis, and real-time secure communications has led to the discovery of many outbreaks that would likely not have been identified with traditional epidemiological methods [9, 22, 23]. Among its successes, PulseNet was instrumental in the recent identification and investigation of an outbreak of E. coli O157:H7 that involved 205 cases from across the United States and Canada [24, 25]. Epidemiologic follow-up of the PulseNet-identified cluster confirmed that 95% of the infected individuals had reported eating fresh, uncooked spinach during the 10 days prior to illness. The association of fresh spinach as a common source was subsequently confirmed by the isolation of the outbreak strain from three open bags of spinach that were recovered from patients’ homes. The rapid and conclusive identification of the contaminated food product and implicated lot numbers resulted in faster and more specific consumer advisories, narrowed the recall of affected food products, and almost certainly limited the scope and severity of the outbreak. The origin of the contaminated spinach was traced to fields in several California counties. PFGE analysis of a cross section of environmental isolates from the San Benito area identified an exact match to the outbreak pattern, isolating the likely point of contamination to within a very narrow geographic margin and improving our understanding of predisposing events [24, 26]. With over a decade of successes such as these, PulseNet has proven to be an important tool for detection, investigation, control, and prevention of foodborne outbreaks within the United States (Table 1). Since its inception, national participation has grown to include all 50 states, as well as county and municipal public health laboratories and food regulatory agencies. In an increasingly globalized economy, the internationalization of PulseNet is a logical progression, but one which requires the ongoing cooperation and commitment of many international partners. The expansion of PulseNet coverage into global markets may also help to enhance food safety within the United States by expediting the identification of contaminated imports before or shortly after they enter the domestic food supply. More importantly, PulseNet International will build upon the successful foundations of PulseNet USA to help us better understand PULSENET INTERNATIONAL 87 the epidemiology and prevention of foodborne infections on a global scale, and enhance food safety and public health throughout the industrialized and developing world.

6.4 PULSENET INTERNATIONAL

Over the past several decades, consumer demand has driven an explosive increase in both the scale and diversity of international food trades [12, 27]. In today’s global market place, foods that are produced, processed, and packaged in one part of the world are more likely than ever to cross national borders, for consumption in countries many hundreds or even thousands of miles away. Foodborne outbreaks may extend well beyond a single region or country, involve multiple imported/exported food products and include, inconsistent recordkeeping, and complex distributorships [6]. These factors can all greatly complicate the identiﬁcation and investigation of international outbreaks, particularly when the scope of most existing foodborne surveillance systems is limited. In order to adapt to these changing parameters, surveillance programs must look beyond national borders to consider the role of economic globalization, and its impact on the safety of domestic food supplies. Internationalization of programs such as PulseNet will help to safeguard food safety within the United States, among our trade partners and within developing nations, where food and waterborne infections continue to represent a signiﬁcant cause of morbidity and mortality [12].

6.4.1 Structure and Function PulseNet International was established as an umbrella organization for independent regional PulseNet networks to provide a cooperative framework for enhanced foodborne disease surveillance on a global scale [12]. Governed by a steering committee that is chaired by the chief of PulseNet USA, PulseNet International currently includes six participating regional networks (USA, Canada, Latin America, Europe, Asia-Pacific, and Middle East) that represent 67 countries across 6 continents (Fig. 6.1). Expansion onto the global stage began with an informal collaboration between PulseNet USA and Canadian public health officials in 1999. This collaboration proved to be useful almost immediately, with the investigation of an outbreak of 91 cases of Salmonella enterica serotype Muenchen in 15 US states and 2 Canadian provinces [28]. Standardized molecular subtyping was able to rapidly identify cases, and suggested an association with the consumption of unpasteurized orange juice, which was later confirmed by epidemiologic follow-up. Cooperation between Canadian and US public health laboratories continued to provide concrete results over the next several years, reinforcing the value of the open and rapid exchange of molecular subtyping information, and its potential—both as an early warning system and as a tool for the investigation of transnational outbreaks [29–34]. The success of this first international collaboration promoted the foundation of PulseNet International. Over the past 5 years, regular organizational meetings have been held between the CDC and public health officials from Canada, Europe, Asia-Pacific, Latin America, and the Middle East to continue expanding the role of PulseNet in global surveillance for foodborne disease. At many of these meetings, the focus has been on establishing necessary infrastructure within each of the regional networks, 88 PULSENET: A PROGRAM TO DETECT AND TRACK FOOD CONTAMINATION EVENTS

PulseNet Canada PulseNet Europe

PulseNet USA PulseNet Asia Pacific PulseNet PulseNet Middle East Latin America

FIGURE 6.1 Delineation of the six PulseNet International regions including PulseNet USA, PulseNet Canada, PulseNet Europe, PulseNet Asia-Pacific, PulseNet Latin America, and PulseNet Middle East [35]. and the integration of each regional network into PulseNet International. Extensive progress has been made with training in PulseNet Standardized PFGE Protocols, software-assisted gel analysis, and establishment of consistent laboratory and analytic QA/QC programs across each regional network [12]. PulseNet International has the added challenge of establishing effective surveillance networks between regional and national partners, whose public health interests may be overshadowed or complicated by important geopolitical differences. The composition and organizational structure within each of the PulseNet International regional networks is flexible, and may differ according to political, economic, and cultural structures within its membership. Despite these challenges, all regional networks have reached consensus on most core operational issues, and identified appropriate means by which to ensure effective, and equitable laboratory-based surveillance programs between their members. The rapid exchange and comparison of PFGE pattern data are critical to surveillance functions, and consequently, information technology infrastructure for database access and communications is as necessary as the development of laboratory resources in the field. Data issues are similarly complex, and difficulties have emerged related to the implementation and maintenance of network infrastructure, and the establishment of multilateral agreements for information management and exchange. An ideal compromise might permit PulseNet members to log on to different PulseNet International servers with limited read-only or volatile access to conduct simple queries. If pattern matches are identified that extend beyond the user’s home network, relevant epidemiological information, such as isolate source, relevant dates, serotype, and PFGE pattern information would be provided, and outbreak coordinators from all affected jurisdictions would be alerted or invited to participate in an investigation. The first step in these processes came with the signature of a Memorandum of Understanding between PulseNet USA and PulseNet Canada in 2005, which formal- ized data-sharing arrangements between the two nations. This agreement granted limited PULSENET INTERNATIONAL 89 access and cross-querying to public health investigators in both countries.and permitted the direct comparison of PFGE data from Shiga toxin-producing E. coli (STEC) and Salmonella. It is hoped that this memorandum will be the first of many agreements, both within and between the other regional networks that make up PulseNet Interna- tional. In the meantime, a secure PulseNet International ListServ has been established to exchange pattern data and communicate related epidemiological information between networks. This approach is more time consuming, since international investigations currently require managers from each regional/national database to manually compare new patterns against their own local pattern database and notify the ListServ if relevant matches are identified.

6.4.2 Food Safety and Bioterrorism: Critical Need for PulseNet International In recent years, an increase in worldwide terrorist activity has raised the profile of biosecurity, and with it, concerns over the vulnerability of international food supplies to acts of intentional biological or chemical contamination. Biological contamination is of particular concern, since pathogenic microorganisms are relatively easy to acquire, and in the case of bacteria, may be manufactured relatively quickly in significant quantities, with only limited budget and expertise [36–38]. Intentional contamination that occurs early in the supply chain, or before the food leaves its country of origin, may have broad international effects, since contaminated lots may be distributed to multiple countries and sold for consumption well before the contamination is detected. Although the importance of biosecurity has increased in many segments of the food production and trade industries, many aspects of international food supply chain remain vulnerable to attack, and naturally occurring outbreaks involving food imports highlight the feasibility and importance of this threat [38]. Historically, only a handful of cases of foodborne disease have been definitively linked to intentional contamination. A retrospective study of outbreak investigations revealed that 44 (4%) out of 1099 involved causative agents with bioterrorism potential, and of those, intentional contamination was considered as a potential explanation in only six [36]. Modern commercial food distribution systems provide an ideal delivery vehicle for widespread terrorist effect, and the threat, or even perceived threat of intentional contamination of the food supply, reinforces the importance of protection and surveillance as national security priorities [36–38]. It is likely that the intentional contamination of a widely distributed food product will initially resemble the emergence of a naturally occuring outbreak of foodborne disease, with a sudden increase in incident cases, and the implication of a well-known foodborne enteric pathogen [36, 38]. Unless the contaminant is of unusual type, character, or distribution, it is also likely that the emergence of an outbreak will first be detected by PulseNet laboratories as a cluster of identical PFGE patterns, and an outbreak investigation will be initiated. The certainty of intentional contamination may not emerge until well into the investigation of the outbreak, unless the event is publicized by complicit terrorist groups, strain(s) is/are of unusual subtype or character, or if the distribution of cases begins to support deliberate and multifocal contamination. PulseNet data will be crucial to determine the nature and extent of the outbreak, and to support criminal investigations by federal and international law enforcement, should they become necessary. 90 PULSENET: A PROGRAM TO DETECT AND TRACK FOOD CONTAMINATION EVENTS

As with any foodborne outbreak, rapid detection and investigation are essential to mounting an effective response against a biological attack, expediting the traceback and removal of contaminated product(s) from the food supply and limiting the scope of the outbreak. Timely identification and response requires sophisticated laboratory-based real-time surveillance that functions at the level of the national population. Sensitiv- ity and detection lag can both be greatly improved, particularly if frontline laboratories have sufficient capacity to complete and submit subtyping results immediately upon receipt. Ensuring that the public health system is ready to deal with intentional contamination events will require enhancement of preparedness and existing public health infrastructure, and as such, the maintenance of state and territory public health resources is critical [36, 37]. The ultimate purpose of any terrorist attack is to induce significant panic within the civilian population, undermine confidence in government, and threaten civil order [38, 39]. Attacks involving mass casualties and weapons of mass destruction are not necessary to achieve these goals. In fact, scenarios involving limited morbidity and mortality, but which impact products and services that civilian populations use or are exposed to frequently during the course of daily life, can increase public anxiety and impose significant strain on public health and primary health care systems, even in the absence of an attack [38]. Among many foodborne outbreaks, unease and economic loss are typical with the identification and recall of a food product. Consumer confidence may be significantly eroded by an outbreak involving a particular food or brand, and the resulting aversion may affect individual producers or entire segments of the market. Thus, a biological attack or even a threat against consumer food supplies may incur significant financial cost, even if the direct impact on civilian populations is, or is expected to be low. To date, the most significant example of intentional and malicious contamination of domestic food supplies occurred in the town of The Dalles, Oregon, in 1984, when followers of Bagwhan Shree Rajneesh contaminated 12 local restaurant salad bars with Salmonella enterica serotype Typhimurium in an effort to manipulate the outcome of local elections. Although there were no fatalities, the attack sickened 751 people, with 45 requiring hospitalization for acute salmonellosis. During the investigation of these cases, the possibility of intentional contamination was considered, but before the availability of molecular surveillance systems such as PulseNet, it took an exhaustive, year long study to link the commune to the outbreak [40]. The simplicity and efficacy of these attacks demonstrate the weaponized potential of foodborne bacteria, and their application to the subversion or manipulation of political processes. In the present day, where terrorist ideals extend beyond the Rajneeshee’s desire to influence local politics, far faster action for response and investigation is required. Among the potential biological agents that the CDC cited in a Strategic Planning Work- group on Biological and Chemical Terrorism were Clostridium botulinum, Salmonella spp., E. coli O157, and V. cholerae [41]. In the United States, PulseNet standardized PFGE protocols and databases are already in place for all of these pathogens with the exception of C. botulinum, although a protocol is currently under development in collaboration with the Virginia Consolidated Laboratory Services. With an established sentinel network that extends throughout North America, and the cooperation of an increasing number of international partners, PulseNet is ideally positioned for the detection and investigation of foodborne bioterrorism, and is highly optimized for the bacterial agents most likely to be involved. RESEARCH DIRECTIONS 91

6.5 RESEARCH DIRECTIONS

Both the present and future success of PulseNet as a laboratory-based surveillance system depend upon its ability to accurately identify, transmit, and interpret bacterial subtyping data. This function must be accomplished rapidly and across a diverse network of national and international laboratories in order to provide timely and actionable information on the type and scope of an emergent outbreak. Although the concept of real-time bacterial subtyping is easy to understand, its implementation, particularly in the context of a broad geographical network, is neither simple nor straightforward. For one, the process is technology-driven: as a technique, PFGE is highly robust and is amenable to implementation in laboratories with a wide range of resources and expertise. In practice, however, it is a labor intensive and meticulous procedure, and even with highly optimized and standardized protocols, PFGE subtyping of bacteria typically takes between 24 and 48 h from pure culture to a completed fingerprint. Furthermore, because only a limited number of samples may be run in parallel, this throughput may be restrictive in time-sensitive outbreak situations [16, 18, 19]. More importantly, although PFGE affords a relatively high degree of discriminatory power and epidemiologic relevance [13, 15, 16], it is not always able to accurately resolve differences between outbreak and sporadic isolates, particularly among highly endemic strain types. As PulseNet databases have grown, database analysts have noticed a remarkably high degree of clonality among seemingly unrelated sporadic isolates (i.e. S. almonella enterica serotype Typhimurium DT104 complex [42], S. enterica serotype Enteritidis phage type 4, 8, and 13a [43], and E. coli O157:H7 [13]). When common patterns are encountered in an outbreak situation, it is often impossible to differentiate between potentially outbreak-related isolates and unrelated sporadic infections by PFGE alone, and the mis-identification of unrelated cases can greatly complicate epidemiologic follow-up. With the increasing feasibility of routine genomic sequencing and high-throughput genetic analysis, the PulseNet methods development laboratory, in collaboration with state (MN, NC, and MA) and federal partners, has begun to develop and implement next-generation subtyping methods to enhance and complement existing PFGE-based protocols. The first of these new methods, multi-locus variable number tandem repeat analysis (MLVA), amplifies a series of short, tandem repeats within the bacterial genome and determines differences in copy number across all loci using capillary gel electrophoresis for high-resolution fragment analysis [44–46]. Interest in the evaluation of MLVA as a subtyping tool first arose after several studies demonstrated its ability to discriminate within highly clonal, PFGE-indistinguishable microorganisms [45, 47, 48], and MLVA has already proven useful in the investigation of outbreaks involving common PFGE patterns [44]. MLVA protocols for E. coli O157:H7, non-O157 STEC, S. enterica serotype Typhimurium, and L. monocytogenes are presently undergoing development or validation in our laboratory, and as of this writing, several protocols have been released for limited use by select PulseNet laboratories [7, 13]. The analysis of single nucleotide polymorphisms (SNPs) is also being evaluated as a next-generation subtyping approach, with active development of panels for rapid E. coli O157:H7 and non-O157 STEC subtyping/characterization [7, 13, 49]. A key advantage of SNP-based subtyping approaches lies in their genomic ubiquity, and their independence from fragment sizing, which negates the need for complex electrophoretic procedures, pattern normalization, and run times. However, selection and validation of SNP target 92 PULSENET: A PROGRAM TO DETECT AND TRACK FOOD CONTAMINATION EVENTS panels are critical, and must combine sufficient coverage to detect genomic change, while retaining sufficient divergency to accurately discriminate between highly clonal bacterial lineages [49]. The development of meaningful interpretation and classification guidelines to describe and classify genetic changes in terms of strain type requires an extensive catalog of historical isolates, both sporadic- and outbreak associated. Occasionally, techniques prove to be too discriminatory: Several of the MLVA loci, for example, have a high degree of variability that may cause pattern instability even over a relatively short period of time [13, 50]. The continued evaluation of these protocols, in conjuction with PFGE and epidemiological data, should assist in the determination of the degree of variability that can be allowed during an outbreak investigation and the development of appropriate classification guidelines. Of greater concern is that many of the variable number tandem repeat (VNTR) sites that are useful for MLVA are highly specific to a given bacterial species or serotype (e.g. E. coli O157:H7 versus non-O157 STEC [13, 44]), and this greatly limits the range of organisms that can be typed, with a corresponding increase in the number of protocols that must be developed, validated, and maintained. An advantage of methodologies such as MVLA and SNP analysis is that many laboratories in the United States already have the necessary equipment available, although training and program funding continue to be problematic in many jurisdictions. These issues are accentuated in developing countries, and it is expected that new technologies will be initially integrated into PulseNet in a tiered approach and used primarily in situations where greater discriminatory power is required [12]. Laboratories that cannot justify an investment in new hardware, staff, and supplies would be encouraged to submit isolates to regional core laboratories, where resources for new technology could be centralized and more effectively supported or funded. Future generations of bacterial subtyping have the potential to be more rapid and less technically demanding, while providing a higher level of epidemiological concordance than PFGE. However, the ultimate utility of these techniques is still being assessed, and their success will depend largely on their amenability to a diverse network of national and international partners. The introduction of any new subtyping technologies to the PulseNet program must be carefully weighed against instrumentation, infrastructural and staffing requirements, the compatibility and reliability of new data relative to existing methods, and by our own capacity to provide training and ongoing support to participating laboratories [12, 13]. With a rapidly expanding network of participating laboratories, PulseNet International is positioned at the forefront of global food safety and biosecurity initiatives. Building upon a history of successes in outbreak detection, response and prevention, it is our hope that the integration of new molecular and informatic technologies will further improve both the foundation and fabric of the PulseNet community, enhancing food safety and public health on a global scale.

REFERENCES

1. Mead, P. S., Slutsker, L., Dietz, V., McCaig, L. F., Bresee, J. S., Shapiro, C., Grifﬁn, P. M., and Tauxe, R. V. (1999). Food-related illness and death in the United States. Emerg. Infect. Dis. 5, 607–625. REFERENCES 93

2. Allos, B. M., Moore, M. R., Griffin, P. M., and Tauxe, R. V. (2004). Surveillance for sporadic foodborne disease in the 21st century: the FoodNet perspective. Clin. Infect. Dis. 38(Suppl 3), S115–S120. 3. Sobel, J., Griffin, P. M., Slutsker, L., Swerdlow, D. L., and Tauxe, R. V. (2002). Investigation of multistate foodborne disease outbreaks. Public Health Rep. 117, 8–19. 4. Tauxe, R. V. (1997). Emerging foodborne diseases: an evolving public health challenge. Emerg. Infect. Dis. 3, 425–434. 5. Tauxe, R. V. (1998). New approaches to surveillance and control of emerging foodborne infectious diseases. Emerg. Infect. Dis. 4, 455–456. 6. Tauxe, R. V., and Hughes, J. M. (1996). International investigation of outbreaks of foodborne disease. BMJ 313, 1093–1094. 7. Ribot, E. R., Hyytia-Trees, E., and Cooper, K. (2007). PulseNet and emerging foodborne diseases. In Microbial Food Contamination, 2nd ed., C. L. Wilson, Ed. CRC Press, Boco Raton, FL, pp. 3–29. 8. Jones, T. F., Scallan, E., and Angulo, F. J. (2007). FoodNet: overview of a decade of achieve- ment. Foodborne Pathog. Dis. 4, 60–66. 9. Tauxe, R. V. (2006). Molecular subtyping and the transformation of public health. Foodborne Pathog. Dis. 3,4–8. 10. Gerner-Smidt, P., Hise, K., Kincaid, J., Hunter, S., Rolando, S., Hyytia-Trees, E., Ribot, E. M., and Swaminathan, B. (2006). PulseNet USA: a five-year update. Foodborne Pathog. Dis. 3, 9–19. 11. Swaminathan, B., Barrett, T. J., Hunter, S. B., and Tauxe, R. V. (2001). PulseNet: the molecular subtyping network for foodborne bacterial disease surveillance, United States. Emerg. Infect. Dis. 7, 382–389. 12. Swaminathan, B., Gerner-Smidt, P., Ng, L. K., Lukinmaa, S., Kam, K. M., Rolando, S., Gutierrez, E. P., and Binsztein, N. (2006). Building PulseNet International: an interconnected system of laboratory networks to facilitate timely public health recognition and response to foodborne disease outbreaks and emerging foodborne diseases. Foodborne Pathog. Dis. 3, 36–50. 13. Hyytia-Trees, E. K., Cooper, K., Ribot, E. M., and Gerner-Smidt, P. (2007). Recent developments and future prospects in subtyping of foodborne bacterial pathogens. Future Microbiol. 2, 175–185. 14. Fakhr, M. K., Nolan, L. K., and Logue, C. M. (2005). Multilocus sequence typing lacks the discriminatory ability of pulsed-field gel electrophoresis for typing Salmonella enterica serovar Typhimurium. J. Clin. Microbiol. 43, 2215–2219. 15. Barrett, T. J., Gerner-Smidt, P., and Swaminathan, B. (2006). Interpretation of pulsed-field gel electrophoresis patterns in foodborne disease investigations and surveillance. Foodborne Pathog. Dis. 3, 20–31. 16. Ribot, E. M., Fair, M. A., Gautom, R., Cameron, D. N., Hunter, S. B., Swaminathan, B., and Barrett, T. J. (2006). Standardization of pulsed-field gel electrophoresis protocols for the subtyping of Escherichia coli O157:H7, Salmonella,andShigella for PulseNet. Foodborne Pathog. Dis. 3, 59–67. 17. Cooper, K. L., Luey, C. K., Bird, M., Terajima, J., Nair, G. B., Kam, K. M., Arakawa, E., Safa, A., Cheung, D. T., Law, C. P., Watanabe, H., Kubota, K., Swaminathan, B., and Ribot, E. M. (2006). Development and validation of a PulseNet standardized pulsed-field gel electrophoresis protocol for subtyping of Vibrio cholerae. Foodborne Pathog. Dis. 3, 51–58. 18. Graves, L. M., and Swaminathan, B. (2001). PulseNet standardized protocol for subtyping Listeria monocytogenes by macrorestriction and pulsed-field gel electrophoresis. Int. J. Food Microbiol. 65, 55–62. 94 PULSENET: A PROGRAM TO DETECT AND TRACK FOOD CONTAMINATION EVENTS

19. Parsons, M. B., Cooper, K. L., Kubota, K. A., Puhr, N., Simington, S., Calimlim, P. S., Schoonmaker-Bopp, D., Bopp, C., Swaminathan, B., Gerner-Smidt, P., and Ribot, E. M. (2007). PulseNet USA standardized pulsed-field gel electrophoresis protocol for subtyping of Vibrio parahaemolyticus. Foodborne Pathog. Dis. 4, 285–292. 20. Ribot, E. M., Fitzgerald, C., Kubota, K., Swaminathan, B., and Barrett, T. J. (2001). Rapid pulsed-field gel electrophoresis protocol for subtyping of Campylobacter jejuni. J. Clin. Micro- biol. 39, 1889–1894. 21. Sealy, T., Kubota, K., and Chu, M. (2003). One-day rapid pulsed-field gel electrophoresis (PFGE) protocol for typing of Yersinia pestis. 103rd General Meeting of the American Society for Microbiology. Washington, DC. 22. Bender, J. B., Hedberg, C. W., Besser, J. M., Boxrud, D. J., MacDonald, K. L., and Osterholm, M. T. (1997). Surveillance by molecular subtype for Escherichia coli O157:H7 infections in Minnesota by molecular subtyping. N. Engl. J. Med. 337, 388–394. 23. Bender, J. B., Hedberg, C. W., Boxrud, D. J., Besser, J. M., Wicklund, J. H., Smith, K. E., and Osterholm, M. T. (2001). Use of molecular subtyping in surveillance for Salmonella enterica serotype typhimurium. N. Engl. J. Med. 344, 189–195. 24. CALFERT (2007). Investigation of an Escherichia coli O157:H7 Outbreak Associated with Dole Pre-Packaged Spinach. 25. CDC (2006). Ongoing multistate outbreak of Escherichia coli serotype O157:H7 infections associated with consumption of fresh spinach—United States, September 2006. MMWR Morb. Mortal. Wkly. Rep. 55, 1045–1046. 26. Cooley, M., Carychao, D., Crawford-Miksza, L., Jay, M. T., Myers, C., Rose, C., Keys, C., Farrar, J., and Mandrell, R. E. (2007). Incidence and tracking of Escherichia coli O157:H7 in a major produce production region in california. PLoS ONE 2, e1159. 27. Werber, D., Dreesman, J., Feil, F., Van Treeck, U., Fell, G., Ethelberg, S., Hauri, A. M., Roggentin, P., Prager, R., Fisher, I. S., Behnke, S. C., Bartelt, E., Weise, E., Ellis, A., Siitonen, A., Andersson, Y., Tschape, H., Kramer, M. H., and Ammon, A. (2005). International outbreak of Salmonella Oranienburg due to German chocolate. BMC Infect. Dis., 5,1. 28. Chan, E. S., Aramini, J., Ciebin, B., Middleton, D., Ahmed, R., Howes, M., Brophy, I., Mentis, I., Jamieson, F., Rodgers, F., Nazarowec-White, M., Pichette, S. C., Farrar, J., Gutierrez, M., Weis, W. J., Lior, L., Ellis, A., and Isaacs, S. (2002). Natural or raw almonds and an outbreak of a rare phage type of Salmonella enteritidis infection. Can. Commun. Dis. Rep. 28, 97–99. 29. CDC (2002). Multistate outbreaks of Salmonella serotype Poona infections associated with eating cantaloupe from Mexico—United States and Canada, 2000–2002. MMWR Morb. Mortal. Wkly. Rep. 51, 1044–1047. 30. CDC (2004). Outbreak of Salmonella serotype Enteritidis infections associated with raw almonds—United States and Canada, 2003–2004. MMWR Morb. Mortal. Wkly. Rep. 53, 484–487. 31. CDC (1999). Outbreaks of Shigella sonnei infection associated with eating fresh parsley— United States and Canada, July–August 1998. MMWR Morb. Mortal. Wkly. Rep. 48, 285–289. 32. Isaacs, S., Aramini, J., Ciebin, B., Farrar, J. A., Ahmed, R., Middleton, D., Chandran, A. U., Harris, L. J., Howes, M., Chan, E., Pichette, A. S., Campbell, K., Gupta, A., Lior, L. Y., Pearce, M., Clark, C., Rodgers, F., Jamieson, F., Brophy, I., and Ellis, A. (2005). An international outbreak of salmonellosis associated with raw almonds contaminated with a rare phage type of Salmonella enteritidis. J. Food Prot. 68, 191–198. 33. Naimi, T. S., Wicklund, J. H., Olsen, S. J., Krause, G., Wells, J. G., Bartkus, J. M., Boxrud, D. J., Sullivan, M., Kassenborg, H., Besser, J. M., Mintz, E. D., Osterholm, M. T., and Hed- berg, C. W. (2003). Concurrent outbreaks of Shigella sonnei and enterotoxigenic Escherichia coli infections associated with parsley: implications for surveillance and control of foodborne illness. J. Food Prot. 66, 535–541. REFERENCES 95

34. Wu, F. M., Doyle, M. P., Beuchat, L. R., Wells, J. G., Mintz, E. D., and Swaminathan, B. (2000). Fate of Shigella sonnei on parsley and methods of disinfection. J. Food Prot. 63, 568–572. 35. El Sedawy, A. (2007), posting date. http://www.pulsenetinternational.org. [Online.] 36. Ashford, D. A., Kaiser, R. M., Bales, M. E., Shutt, K., Patrawalla, A., McShan, A., Tappero, J. W., Perkins, B. A., and Dannenberg, A. L. (2003). Planning against biological terrorism: lessons from outbreak investigations. Emerg. Infect. Dis. 9, 515–519. 37. Sandhu, H. S., Thomas, C., Nsubuga, P., and White, M. E. (2003). A global network for early warning and response to infectious diseases and bioterrorism: applied epidemiology and training programs, 2001. Am. J. Public Health 93, 1640–1642. 38. Sobel, J., Khan, A. S., and Swerdlow, D. L. (2002). Threat of a biological terrorist attack on the US food supply: the CDC perspective. Lancet 359, 874–880. 39. Polyak, C. S., Macy, J. T., Irizarry-De La Cruz, M., Lai, J. E., McAuliffe, J. F., Popovic, T., Pillai, S. P., and Mintz, E. D. (2002). Bioterrorism-related anthrax: international response by the Centers for Disease Control and Prevention. Emerg. Infect. Dis. 8, 1056–1059. 40. Torok, T. J., Tauxe, R. V., Wise, R. P., Livengood, J. R., Sokolow, R., Mauvais, S., Birkness, K. A., Skeels, M. R., Horan, J. M., and Foster, L. R. (1997). A large community outbreak of salmonellosis caused by intentional contamination of restaurant salad bars. JAMA 278, 389–395. 41. CDC (2000). Biological and chemical terrorism: strategic plan for preparedness and response. Recommendations of the CDC Strategic Planning Workgroup. MMWR Recomm. Rep. 49, 1–14. 42. Liebana, E., Garcia-Migura, L., Clouting, C., Clifton-Hadley, F. A., Lindsay, E., Threlfall, E. J., McDowell, S. W., and Davies, R. H. (2002). Multiple genetic typing of Salmonella enterica serotype typhimurium isolates of different phage types (DT104, U302, DT204b, and DT49) from animals and humans in England, Wales, and Northern Ireland. J. Clin. Microbiol. 40, 4450–4456. 43. Ridley, A. M., Threlfall, E. J., and Rowe, B. (1998). Genotypic characterization of Salmonella enteritidis phage types by plasmid analysis, ribotyping, and pulsed-ﬁeld gel electrophoresis. J. Clin. Microbiol. 36, 2314–2321. 44. Hyytia-Trees, E., Smole, S. C., Fields, P. A., Swaminathan, B., and Ribot, E. M. (2006). Second generation subtyping: a proposed PulseNet protocol for multiple-locus variable-number tandem repeat analysis of Shiga toxin-producing Escherichia coli O157 (STEC O157). Foodborne Pathog. Dis. 3, 118–131. 45. Keim, P., Price, L. B., Klevytska, A. M., Smith, K. L., Schupp, J. M., Okinaka, R., Jackson, P. J., and Hugh-Jones, M. E. (2000). Multiple-locus variable-number tandem repeat analysis reveals genetic relationships within Bacillus anthracis. J. Bacteriol. 182, 2928–2936. 46. van Belkum, A., Scherer, S., van Alphen, L., and Verbrugh, H. (1998). Short-sequence DNA repeats in prokaryotic genomes, Vol. 62, p. 275–293. 47. Keys, C., Kemper, S., and Keim, P. (2005). Highly diverse variable number tandem repeat loci in the E. coli O157:H7 and O55:H7 genomes for high-resolution molecular typing. J. Appl. Microbiol. 98, 928–940. 48. Lindstedt, B. A., Vardund, T., Aas, L., and Kapperud, G. (2004). Multiple-locus variable-number tandem-repeats analysis of Salmonella enterica subsp. enterica serovar Typhimurium using PCR multiplexing and multicolor capillary electrophoresis. J. Microbiol. Methods 59, 163–172. 49. Zhang, W., Qi, W., Albert, T. J., Motiwala, A. S., Alland, D., Hyytia-Trees, E. K., Ribot, E. M., Fields, P. I., Whittam, T. S., and Swaminathan, B. (2006). Probing genomic diversity and evolution of Escherichia coli O157 by single nucleotide polymorphisms. Genome Res. 16, 757–767. 96 PULSENET: A PROGRAM TO DETECT AND TRACK FOOD CONTAMINATION EVENTS

50. Noller, A. C., McEllistrem, M. C., Shutt, K. A., and Harrison, L. H. (2006). Locus-speciﬁc mutational events in a multilocus variable-number tandem repeat analysis of Escherichia coli O157:H7. J. Clin. Microbiol. 44, 374–377. 51. Mahon, B. E., Ponka, A., Hall, W. N., Komatsu, K., Dietrich, S. E., Siitonen, A., Cage, G., Hayes, P. S., Lambert-Fair, M. A., Bean, N. H., Grifﬁn, P. M., and Slutsker, L. (1997). An international outbreak of Salmonella infections caused by alfalfa sprouts grown from contaminated seeds. J Infect Dis. 175, 876–882. 52. MMWR. (2005). Escherichia coli O157:H7 infections associated with ground beef from a U.S. military installation–Okinawa, Japan, February 2004.MMWR Morb. Mortal. Wkly. Rep. 54, 40–42.

FURTHER READING

Gerner-Smidt, P., Hise, K., Kincaid, J., Hunter, S., Rolando, S., Hyytia-Trees, E., Ribot, E. M., and Swaminathan, B. (2006). PulseNet USA: a ﬁve-year update. Foodborne Pathog. Dis. 3, 9–19. Hyytia-Trees, E. K., Cooper, K., Ribot, E. M., and Gerner-Smidt, P. (2007). Recent developments and future prospects in subtyping of foodborne bacterial pathogens. Future Microbiol. 2, 175–85. Sobel, J., A. S. Khan, and Swerdlow, D. L. (2002). Threat of a biological terrorist attack on the US food supply: the CDC perspective. Lancet 359, 874–80. Swaminathan, B., Barrett, T. J., Hunter, S. B., and Tauxe, R. V. (2001). PulseNet: the molecular subtyping network for foodborne bacterial disease surveillance, United States. Emerg. Infect. Dis. 7, 382–9. Swaminathan, B., Gerner-Smidt, P., Ng, L. K., Lukinmaa, S., Kam, K. M., Rolando, S., Gutierrez, E. P., and Binsztein, N. (2006). Building PulseNet International: an interconnected system of laboratory networks to facilitate timely public health recognition and response to foodborne disease outbreaks and emerging foodborne diseases. Foodborne Pathog. Dis. 3, 36–50. Voeller V05-c07.tex V1 - 12/04/2013 1:06pm Page 97

7 INSECTS AS VECTORS OF FOODBORNE PATHOGENS

Ludek Zurek Kansas State University, Departments of Entomology and Diagnostic Medicine and Pathobiology, Manhattan, Kansas J. Richard Gorham United States Public Health Service, Food and Drug Administration, Xenia, Ohio

7.1 INTRODUCTION

Two areas of concern are discussed in this chapter. One, the major one, has to do with the contamination of food and food-contact surfaces by various insect pests often associated with human or animal foods [1]. The scenarios by which such contaminations occur are well known and are mitigated by strict adherence to sanitation standard operating procedures (SSOPs) and good manufacturing practices (GMPs), by the implementation of the hazard analysis critical control points (HACCP) program, and by the practice of Integrated Pest Management (IPM). We will not describe these four programs. The reader will ﬁnd abundant resources about these programs on the Internet, from the Land Grant universities, scientiﬁc literature, and commercial providers of these programs [2, 3]. The lesser concern, a much less familiar one, deals with intentional food contamination mediated by insect agents. To deal with this threat, an equally proactive approach, similar to SSOPs/GMPs/HACCP/IPM, is essential. It involves a strategy we have termed AIM=F: anticipate, inform, mitigate equals frustrate, that is, the prevention, neutralization or control of intentional acts of food contamination by means of insect agents.

7.2 MUSCOID FLIES AND FRUIT FLIES

Muscoid flies and fruit flies represent a close association of insects with microbes, especially with bacteria originating from human and animal feces and other decaying organic materials. Moreover, muscoid flies have a great potential to contaminate human food and drink with bacteria, including foodborne pathogens, because of their developmental

97 Voeller V05-c07.tex V1 - 12/04/2013 1:06pm Page 98

98 INSECTS AS VECTORS OF FOODBORNE PATHOGENS habitats, mode of feeding (regurgitation), unrestricted movement, and attraction to places occupied by humans and domestic animals.

7.2.1 Nutrition and Development Virtually any environment rich in decaying organic matter harbors a diverse bacterial community and becomes a suitable substrate for development of muscoid flies, such as house flies (Musca domestica), stable flies (Stomoxys calcitrans), horn flies (Haematobia irritans), and face flies (Musca autumnalis) [4]. The primary larval developmental sites for these flies include animal feces/manure and other decaying organic material (human garbage and compost). The importance of bacteria in the development of muscoid flies has been reported in several studies that show that a live bacterial community is essential for the larval development of these flies. The nature of this symbiosis is unclear. The significance of bacteria for the development of larvae has been examined for house flies [5–7], stable flies [8, 9], horn flies [10], and face flies [11]. Digestibility of bacteria in the intestinal tract was demonstrated in house flies [12], stable flies [13], and blow flies [14, 15]. Other studies of morphological and physiological adaptations of muscoid flies for uptake, storage, and digestion of bacteria also emphasized the importance of bacteria in larval development [12, 16]. In addition, it has been demonstrated that the same bacteria that support the development of stable fly larvae also stimulate oviposition (egg laying) on the specific substrate and therefore indicate the suitability of the substrate for offspring development [9]. Studies on house flies and stable flies have demonstrated that bacteria in the larval gut can survive pupation and can colonize the digestive tract of newly emerged adult flies [17, 18]. This important finding supports the idea that adult muscoid flies serve as vectors of human and animal pathogenic bacterial strains. Fruit flies do not require bacteria to successfully complete development; however, it has been shown that exogenous bacteria enhance the lifespan of Drosophila melanogaster, especially during the first week of adult life [19]; however, a more recent study did not confirm these results [20].

7.2.2 Dissemination of Pathogens and Antibiotic Resistant Strains House flies and other muscoid (filth) flies are pests of great medical and veterinary significance [21]. House flies are important nuisance pests of domestic animals and people, as well as the main fly vectors of foodborne and animal pathogens [21–23]. Due to their indiscriminate movements, ability to fly long distances, and attraction to both decaying organic materials and places where food is prepared and stored, house flies greatly amplify the risk of human exposure to foodborne pathogens. House flies can transport microbial pathogens from reservoirs (animal manure) where they present a minimal hazard to people to places where they pose a great risk (food) [21, 22]. Stable flies are bloodsucking insects and important pests of domestic animals and people. Stable flies cause great economic losses in the animal industry, primarily in dairy and beef production [24, 25], and they can also play a role in ecology of various bacteria originating from animal manure and other larval developmental habitats [18]. The potential of adult house flies to transmit pathogens such as Yersinia pseudotuberculosis [26, 27], Helicobacter pylori [28], Campylobacter jejuni [29], Escherichia coli O157:H7 [30–32], Salmonella spp. [33], and Aeromonas caviae [34] has been also reported. Recently, it has been demonstrated Voeller V05-c07.tex V1 - 12/04/2013 1:06pm Page 99

MUSCOID FLIES AND FRUIT FLIES 99 that house flies are capable of transmitting E. coli O157:H7 to cattle, the major reservoir of this human foodborne pathogen [35]. Fruit flies, primarily the Mediterranean fruit fly (Ceratitis capitata) and the vinegar fruit fly (D. melanogaster), were also reported as potentially competent vectors for E. coli O157:H7 and were capable of contaminating fruits with this pathogen under laboratory conditions [36, 37]. Several studies reported a direct positive correlation between the incidence of foodborne diarrheal diseases and the density of fly populations. For example, suppression of flies in a military camp in the Persian Gulf region resulted in an 85% decrease in shigellosis and a 42% reduction in the incidence of other diarrheal diseases [38]. Esrey [39] reported a 40% reduction of incidence of diarrheal infections in children after suppression of the fly population. Additionally, the development of antibiotic resistance among clinical bacterial isolates and commensal bacteria of people and animals, as well as bacteria in other habitats, raises a concern that flies may be vector competent not only for specific pathogens but also for nonpathogenic bacteria carrying antibiotic resistance genes. A recent study reported that the majority of house flies collected from fast-food restaurants in the United States carried a large population of antibiotic resistant and potentially virulent Enterococci, primarily Enterococcus faecalis. The resistance genes were present on mobile genetic elements (plasmids, transposons) with a broad host range [40] that could be potentially transferred by horizontal gene transfer to more pathogenic strains. Additionally, it has been shown that ready-to-eat food in fast-food restaurants is more frequently contaminated by E. faecalis and Enterococcus faecium in summer months when house flies are more common in restaurants than in winter months [41], indirectly implicating house flies as a potential source of the contamination.

7.2.3 Homeland Security Aspects It is becoming more apparent that muscoid flies, primarily house flies, and some species of fruit flies have the potential to play an important role in the dissemination of foodborne pathogens in both agricultural and urban environments. Consequently, both preharvest and postharvest food safety strategies will have to include the insect pest management approach. Unfortunately, the current mind set of many farmers and animal production managers is to tolerate insects such as house flies (and other pests that do not have direct and obvious economic impact on animal production) unless residents from surrounding urban sites complain about fly or other insect infestation problems. House flies and fruit flies can be easily reared in large numbers in laboratory colonies and could be intentionally contaminated on the surface and in the digestive tract by various bacteria, including foodborne pathogens such E. coli O157:H7, Salmonella spp., and Campylobacter spp. Although muscoid flies and fruit flies have been shown to carry these bacteria in nature and have potential to contaminate the surfaces and food they feed on, the relatively short life span of these flies (up to 2–3 weeks) probably does not represent a viable prospect for domestic or international bioterrorist attack that would have serious consequences on a large scale. However, the AIM = F (anticipate, inform, mitigate equals frustrate) strategy has to be ready for this scenario because the typical integrated pest management (lPM) approach would be too slow to protect the public. Immediate quarantine and insecticide measures will have to be in place and ready to be implemented for such situations. Voeller V05-c07.tex V1 - 12/04/2013 1:06pm Page 100

100 INSECTS AS VECTORS OF FOODBORNE PATHOGENS

7.3 COCKROACHES

7.3.1 Nutrition and Development Cockroaches (Blattaria, Dictyoptera) of many species are widely distributed in the natural world, but only a relatively few species have adapted to life within manmade structures or to the habit of frequently invading such structures from the outdoors [42, 43]. Foraging for food generally occurs at night. Cockroaches typically retire to dark, sheltered niches during the hours of daylight. Gradual metamorphosis being the rule in the Blattaria, nymphs emerging from eggs lack wings and functional reproductive organs, but otherwise they are similar to the adult stage except for being smaller in size. All postegg stages have chewing mouthparts and all utilize similar kinds of food. They are omnivores; virtually any organic material, of either plant or animal origin and either solid or liquid, can be ingested. Domestic cockroaches tend to require a daily ration of water. This may be supplied as liquid water, as in a ﬂoor drain or a puddle under leaky plumbing, or in the form of moist food (anything from food on a hospital food cart to rotting kitchen waste in a garbage can). Moisture, as well as food, may be acquired by ingesting human or animal feces, vomitus, blood, and pus on discarded wound dressings, and moist pet food, to name a few sources. When it comes to food and drink, cockroaches take whatever they can get wherever they can get it. This is where the problem arises for human and animal health: Like ﬂies, cockroaches visit feces (and many other contaminated substrates) and food (that is, edible human or animal food) indiscriminately and their movements from one to the other may contaminate food-contact surfaces.

7.3.2 Dissemination of Pathogens The cockroach gut is home to a bewildering array of naturally occurring bacteria, most of which are harmless to people and domestic animals [44–46]. But in their visits to substrates laden with pathogens, their exterior surfaces, especially the legs, become laden with pathogenic bacteria. Moreover, they can ingest pathogens, some of which may survive in the gut long enough to be egested with the fecal pellets or, occasionally, regurgitated during feeding. Thus, both clean surfaces and clean food may become contaminated. Although some doubt about the importance of cockroaches as vectors of foodborne pathogens has been expressed [47], the larger body of published research, some of which is noted here, suggests that cockroaches should be given serious consideration by the public and by the guardians of the public’s health. Concern over the role of flies, cockroaches, and ants as potential vectors of microbes pathogenic to humans and animals dates at least from very early in the 1900s and this concern is reflected in the many dozens of scientific papers published during the past century. There is much to be learned from these older papers; many of them are cited in more recent papers and several of them are appended in “Further Reading”. For this section on cockroaches, we will bring to the reader’s attention a few investigative reports published since the turn of the present century. The essential thrust of these papers is that pathogens and cockroaches are intimately and consistently associated, a conclusion derived from multiple isolations of pathogens from cockroaches collected in places, such as hospitals and kitchens, generally perceived to be sanitary and sanitized. Cockroaches and their associated pathogens might be implicated in some way, either by direct contact with people (or domestic animals), or by contact with food or food-contact surfaces, is a premise supported by the observations Voeller V05-c07.tex V1 - 12/04/2013 1:06pm Page 101

COCKROACHES 101 that specific disease outbreaks waned when standard infection control procedures were complemented by elimination of cockroaches [48, 49]. None of these reports conclusively proves that the cockroach committed the “crime,” but the correlation of the specific strain of the pathogen taken from the cockroach with the same specific strain taken from the sick patient seems to us to be very compelling circumstantial evidence implicating the cockroach. The authors of virtually every scientific paper on this subject published since 1900 have come to this understanding. Two other factors add weight to the premise that cockroaches and food (or food-contact surfaces) should not coincide: (i) some strains of pathogens exhibit enhanced virulence, that is, even an immunologically competent host may be susceptible to a much lower than usual infective dose; and (ii) immunocompromised hosts are, of course, susceptible to the supervirulent strains and to lower than usual infective doses of the standard pathogenic strains. All agree that the cornerstone of personal and community hygiene is hand-washing. Countless incidents of foodborne disease and nosocomial infections have been traced back to a simple behavioral flaw: hand-washing was omitted or done ineffectively. People can be trained to more consistently and effectively wash their hands. Although flies, ants and cockroaches engage in a lot of self-grooming, a behavior vaguely comparable to hand-washing, this does not render them clean in the microbiological sense, as has been graphically demonstrated in at least one instance for cockroaches [50]. We offer here a partial list of pathogens isolated from various species of common domestic cockroaches (locality information, given only after first mention of a given reference, is stated after the reference number); many other pathogen isolation reports may be found in the extensive literature on this subject [51]. Although the status of each of the several pathogens with regard to antibiotic resistance, a very common phenomenon, may be of special interest to clinicians, this information is omitted here because the matter does not seem essential to the purposes of this chapter. Aeromonas [52 (Libya); 53 (Nigeria)]; Bacillus sp. [54 (Botswana)]; Citrobacter freundii [53, 55 (Thailand)]; Enterobacter aerogenes [56 (Brazil)]; Enterobacter cloacae [53, 55, 56]]; Enterobacter gergoviae [56]; Enterobacter sp. [52, 54]; Erwinia sp. [54]; E. coli [[53–55] 57 (Taiwan)]: Hafnia alvei [56]; Klebsiella pneumoniae [48 (South Africa); [53, 55, 56]]; Klebsiella sp. [52, 54]; Mycobacteria [58 (Taiwan)]; Proteus mirabilis [53]; Proteus sp. [57]; Proteus vulgaris [53]; Pseudomonas aeruginosa [53, 57]; Pseudomonas sp. [54]; Salmonella sp. [53, 54]; Serratia marcescens [53, 56, 57]; Serra- tia sp. [52, 54, 56]; Shigella sp. [54]; Staphylococci (Gram neg.) [56]; Staphylococcus aureus [53, 57]; Staphylococcus epidermidis [53]; Staphylococcus sp. [54]; Streptococcus faecalis [53]; Streptococcus sp. [52]; Alternaria sp. [59 (Brazil)]; Aspergillus flavus [54]; Aspergillus fumigatus [54]; Aspergillus parasiticus [54]; Aspergillus sp. [59]; Candida sp. [53, 54, 59]; filamentous fungi [56]; Penicillium sp. [59]; yeast [56]; Ballantidium coli [53]; Cryptosporidium parvum [53]; Entamoeba histolytica [60 (Taiwan)]; Ancylostoma duodenale [53]; Ascaris lumbricoides [53]; Enterobius vermicularis [53]; Strongyloides stercoralis [53]; Trichuris trichiura [53].

7.3.3 Homeland Security Aspects Our primary concern here is to keep our citizens healthy and productive by ensuring that their food is safe to eat. One of the many ways to do that is to prevent the convergence of food and cockroaches, a convergence that is still much too common. Voeller V05-c07.tex V1 - 12/04/2013 1:06pm Page 102

102 INSECTS AS VECTORS OF FOODBORNE PATHOGENS

Several species of domestic cockroaches, especially Blattella germanica (Blattelli- dae), Blatta orientalis (Blattidae), and Periplaneta americana (Blattidae), can be easily reared in huge numbers in the laboratory and are easily contaminated, either superficially or internally, with certain pathogens (such as avian influenza virus, SARS virus, foot-and-mouth disease virus, E. coli O157:H7, to name a few) that may cause disease in humans or in domestic animals (and then, in the latter case, may secondarily cause disease in humans). Cockroaches, upon their release from the rearing environment, typically first seek shelter. As the light of day wanes, the cockroaches will venture forth in search for moisture. Some fall into and drown in the water supplies that serve the chickens, cows, or pigs, inadvertently releasing their burden of pathogens. Others are eaten by pigs or chickens or accidentally ingested by cows as they feed nose-to-nose with the cockroaches. Others seek out the darkness and moisture of the beverage and ice machines in the school, restaurant or company cafeteria. Again, pathogens are deposited on surfaces presumed to be clean. Whether this shotgun type of dissemination will result in human or animal disease, no one can predict. But the level of probability for that eventuality seems to be at least somewhat higher than what might occur during the normal course of farm and food service operations. Now is the time for the AIM = F strategy to pay off. Thanks to the “A,” our farmers, ranchers, factory managers, food service personnel, and school administrators are aware of the inventory of unfriendly interventions that might occur; they have been “I” (Informed) on how to recognize the signs of enemy interventions; they know that IPM is an effective form of “M” (Mitigation); and the combination of AIM results in the “F” (Frustration) of this assault on the public’s health. In the bioterrorism scenario, it may not be feasible to wait for the slower pest control measures that are typical of the usual lPM approach. Immediate and thorough application of insecticides and immediate quarantine measures may be essential to quell an obvious threat; protocols for these interventions should be in place, practiced and ready for implementation.

7.4 ANTS

7.4.1 Nutrition and Development Ants (Formicidae, Hymenoptera) are social insects, that is, they live in colonies, each colony responding to the control of (usually) only one queen. The worker ants are females. They are the ones that leave the nest and venture out on food-ﬁnding expeditions. Colony size varies greatly according to species and within species. Some are enormous, with thousands of workers; others, only a few dozen. Unlike the cockroaches, ants go through a complete metamorphosis—egg, larva, pupa, adult; but like cockroaches, most kinds of ants live in the natural world; only a relatively few species either nest in manmade structures or routinely forage within such structures [61, 62]. Structure-invading ants are omnivores. The animal proteins and fats in their diet are derived mostly from insects and other arthropods that fall prey to the foraging worker ants. Sugars and starches or foods containing those carbohydrates are often very attractive to ants. Kitchens, bakeries, restaurants, and food factories are typical venues where ants collect a variety of foods that are then held in their chewing mouthparts and transported to the home nest to become essential nutrients for the queen and her brood of larvae. Hospitals too, are often visited. Besides the usual ﬂoor feasts of bread crumbs, sugar granules, and fat droplets, ants, especially the pharaoh ant, Monomorium pharaonis, Voeller V05-c07.tex V1 - 12/04/2013 1:06pm Page 103

PANTRY PESTS 103 may annoy patients by nibbling on food around a patient’s mouth; they also feed on exposed pus and dried blood, or they may be found on patient food trays. These ants (M. pharaonis) have been found in IV drips and inside packages of sterile dressings [63, 64]. Water is essential and this may be obtained from any exposed source such as ﬂoor drains, urinals, patient water ﬂasks, unemptied bedpans, wound dressings, ice machines, plumbing drips, and so forth.

7.4.2 Dissemination of Pathogens Like cockroaches, ants harbor many kinds of internal bacteria [65, 66], but, with a few exceptions, only the external surfaces, mainly the legs and mandibles, are of concern here [1, 67–69]. These appendages come into contact with substrates, such as the soil and pit latrines outdoors and, most commonly, floors indoors, from which the ants may pick up pathogens. As the ants forage over clean surfaces, such as dishes or cutting boards, or food conveyors in a factory, pathogens may be deposited and eventually become mixed in with a food destined, without a subsequent heat treatment, for human or animal consumption. Ants as pests in hospitals have been reported many times [70–74]. We offer here a partial list of pathogens isolated from various species of common pest ants (locality information, given only after first mention of a given reference, is stated after the reference number); other pathogen isolation reports may be found in the literature on this subject. Bacillus cereus [70 (England)]; bacteria (Gram +) [72 (Brazil)]; Clostridium perfringens [70]; E. coli [70]; filamentous fungi [72]; K. pneumoniae [71 (Trinidad)]; Micrococcus sp. [72]; P. mirabilis [71]; Pseudomonas sp. [71]; Salmonella sp. [70]; S. aureus [70]; Staphylococcus sp. [72]; Streptococcus pyogenes [70].

7.4.3 Homeland Security Aspects Although ants are good candidates for the role of accidental mechanical vectors of pathogens, they are poor candidates as pawns in an act of intentional food contamination. The principal homeland security concern here coincides with the universal objective of operating hospitals and food service facilities, including the home kitchen, in such a sanitized manner that food offered for human consumption is safe to eat, that is, at least it and the surfaces it has touched have been protected from exposure to the pathogens that ants and cockroaches are known to carry.

7.5 PANTRY PESTS

7.5.1 Nutrition and Development The moths (Lepidoptera) and beetles (Coleoptera) that infest grains, ﬂour, nuts, chocolate, dry dog food, and cereals in the kitchen storage cabinet are referred to as pantry pests. They are found in home kitchens, of course, but also in grain storage elevators, huge ships that transport grains, bakeries, restaurants, chicken ranches, dairy barns, food factories, food warehouses, transport trucks, and many other venues both large and small. The pantry pests noted here are holometabolous, that is, their life stages are egg, larva, pupa, and adult. The larva has chewing mouthparts; it is the stage that does the bulk of the feeding and the bulk of the damage to commodities. Voeller V05-c07.tex V1 - 12/04/2013 1:06pm Page 104

104 INSECTS AS VECTORS OF FOODBORNE PATHOGENS

7.5.2 Dissemination of Pathogens Compared to ants and cockroaches, pantry pests are relatively free of pathogens that cause human or animal diseases. They do not usually get out into those venues where bacterial pathogens are common. Unfortunately, they often do not long remain free of pathogens [75] or spoilage organisms [76]. This is because their food sources, in which they live throughout their entire lives, are visited by those pests that commonly visit pathogen-laden substrates. Cockroaches, ants, ﬂies, rats, and mice bring pathogens to the home territory of the pantry pests. The latter, then, quite inadvertently spread these pathogens here and there as they move about within their food material [77]. The situation is quite different with regard to spoilage molds. The spores of these fungi are ubiquitous; they are produced most abundantly from grain substrates that are damp and deteriorating, that is, “out of condition.” Grain spoilage represents economic loss; that explains why managers of grain storages, whether for bulk commodities or retail packages, go to great lengths to maintain a dry environment for these products. But beyond the economic consideration, moldy grain can become a health hazard for both people and domestic animals when certain fungi of deterioration produce aﬂatoxins.

7.5.3 Homeland Security Aspects Our concerns here are similar to those faced with ants. The primary goal is to keep susceptible products—nuts, grains, beans, coffee beans, peanuts, and so forth—free of pantry pests, the objective being to produce end-product foods that are safe for human and animal consumption. Generally speaking, the better the storage conditions, the less likely that pantry pests will become established and the less likely that spoilage molds and aflatoxin-producing fungi will proliferate in the commodity. Pantry pests spread the spores of the aflatoxin-producing fungi [78, 79] through the commodity just as they do the spores of common spoilage molds. Several kinds of pest beetles are easy to cultivate in very large numbers. It would be a simple matter to superficially contaminate adult beetles with some pathogen and release them at a vulnerable location. The sudden increase in the population of a pest around or within a food facility would be the signal to implement AIM = F, with emphasis on immediate, focused insecticidal treatment of the affected facility.

REFERENCES

1. Gorham, J. R. (1991). Food pests as disease vectors. In Ecology and Management of Food-industry Pests, FDA Tech Bull 4, J. R. Gorham, Ed. AOAC International, Arlington, VA, pp. 477–482. 2. Hui, Y. H., Nip, W.-K., and Gorham, J. R. (2003). Sanitation and warehousing. In Food Plant Sanitation, Y. H. Hui, B. L. Bruinsma, J. R. Gorham, W.-K. Nip, P. S. Tong, and P. Ventresca, Eds. Marcel Dekker, New York, pp. 373–389. 3. Stanﬁeld, P. (2006). FDA’s GMPs, HACCP, and the food code. In Handbook of Food Science, Technology, and Engineering,Y.H.Hui,Ed.Vol.2, CRC Taylor & Francis, Boca Raton, FL, pp. 73.1–73.14. 4. Spiller, D. (1964). Nutrition and diet of muscoid ﬂies. Bull. World Health Organ. 341, 551–554. Voeller V05-c07.tex V1 - 12/04/2013 1:06pm Page 105

REFERENCES 105

5. Schmidtmann, E. T., and Martin, P. A. W. (1992). Relationship between selected bacteria and the growth of immature house flies, Musca domestica, in an axenic test system. J. Med. Entomol. 29, 232–235. 6. Watson, D. W., Martin, P. A. W., and Schmidtmann, E. T. (1993). Egg yolk and bacteria growth medium for Musca domestica (Diptera: Muscidae). J. Med. Entomol. 30, 820–823. 7. Zurek, L., Schal, C., and Watson, D. W. (2000). Diversity and contribution of the intestinal bacterial community to the development of Musca domestica (Diptera: Muscidae) larvae. J. Med. Entomol. 37(6), 924–928. 8. Lysyk, T. J., Kalischuk-Tymensen, L., Selinger, L. B., Lancaster, R. C., Wever, L., and Cheng, K.-J. (1999). Rearing stable flies larvae (Diptera: Muscidae) on an egg yolk medium. J. Med. Entomol. 36, 382–388. 9. Romero, A., Broce, A., and Zurek, L. (2006). Role of bacteria in the oviposition behavior and larval development of stable flies. Med. Vet. Entomol. 20(1), 115–121. 10. Perotti, M. A., Lysyk, T. J., Kalischuk-Tymensen, L. D., Yanke, L. J., and Selinger, L. B. (2001). Growth and survival of immature Haematobia irritans (Diptera: Mus- cidae) is influenced by bacteria isolated from cattle manure and conspecific larvae. J. Med. Entomol. 38(2), 180–187. 11. Hollis, J. H., Knapp, F. W., and Dawson, K. A. (1985). Influence of bacteria within bovine feces on the development of the face fly (Diptera: Muscidae). Environ. Entomol. 14, 568–571. 12. Espinosa-Fuentes, F. P., and Terra, W. R. (1987). Physiological adaptations for digestion bacteria. Water fluxes and distribution of digestive enzymes in Musca domestica larval midgut. Insect. Biochem. 17, 809–817. 13. Rochon, K., Lysyk, T. J., and Selinger, L. B. (2004). Persistence of Escherichia coli in immature house fly and stable fly (Diptera: Muscidae) in relation to larval growth and survival. J. Med. Entomol. 41(6), 1082–1089. 14. Greenberg, B. (1968). Model for destruction of bacteria in the midgut of blow fly maggots. J. Med. Entomol. 5, 31–38. 15. Mumcuoglu, K. Y., Miller, J., Mumcuoglu, M., Friger, M., and Tarshis, M. (2001). Destruc- tion of bacteria in the digestive tract of the maggot of Lucilia sericata (Diptera: Calliphoridae). J. Med. Entomol. 38(2), 161–166. 16. Dowding, V. M. (1967). The function and ecological significance of the pharyngeal ridges occurring in the larvae of some cyclorrhaphous Diptera. Parasitology 57, 371–388. 17. Greenberg, B. (1959). Persistence of bacteria in the developmental stages of the housefly.4. Infectivity of the newly emerged adult. Am. J. Trop. Med. Hyg. 8(6), 618–622. 18. Rochon, K., Lysyk, T. J., and Selinger, L. B. (2005). Retention of Escherichia coli by house fly and stable fly (Diptera: Muscidae) during pupal metamorphosis and eclosion. J. Med. Entomol. 42(3), 397–403. 19. Brummel, T., Ching, A., Seroude, L., Simon, A. F., and Benzer, S. (2004). Drosophila lifespan enhancement by exogenous bacteria. Proc. Natl. Acad. Sci. U.S.A. 101(35), 12974–12979. 20. Ren, C., Webster, P., Finkel, S. E., and Tower, J. (2007). Increased internal and external bacterial load during Drosophila aging without life-span trade-off. Cell Metab. 6(2), 144–152. 21. Olsen, A. R. (1998). Regulatory action criteria for filth and other extraneous materials III. Review of flies and foodborne enteric disease. Regul. Toxicol. Pharm. 28(3), 199–211. 22. Greenberg, B. (1971). Flies and Diseases, Princeton University Press, Princeton, NJ. 23. Graczyk, T. K., Knight, R., Gilman, R. H., and Cranfield, M. R. (2001). The role of non-biting flies in the epidemiology of human infectious diseases. Microbes Infect. 3(3), 231–235. 24. Campbell, J. B., Berry, I. L., Boxler, D. J., Davis, R. L., Clanton, D. C., and Deutscher, G. H. (1987). Effects of stable flies (Diptera: Muscidae) on weight gain and feed efficiency of feedlot cattle. J. Econ. Entomol. 80, 117–119. Voeller V05-c07.tex V1 - 12/04/2013 1:06pm Page 106

106 INSECTS AS VECTORS OF FOODBORNE PATHOGENS

25. Campbell, J. B., Skoda, S. R., Berkebile, D. R., Boxler, D. J., Thomas, G. D., Adams, D. C., and Davis, R. (2001). Effects of stable flies (Diptera: Muscidae) on weight gains of grazing yearling cattle. J. Econ. Entomol. 94(3), 780–783. 26. Fukushima, H., Tsubokura, M., Otsuki, K., and Kawaoka, Y. (1984). Biochemical hetero- geneity of serotype 03 strains of 700 Yersinia strains isolated from humans, other mammals, flies, animal feed, and river water. Curr. Microbiol. 11, 149–154. 27. Zurek, L., Denning, S. S., Schal, C., and Watson, D. W. (2001). Vector competence of Musca domestica (Diptera: Muscidae) for Yersinia pseudotuberculosis. J. Med. Entomol. 38(2), 333–335. 28. Grubel, P., Hoffman, J. S., Chong, F. K., Burstein, N. E., Mepani, C., and Cave, D. R. (1997). Vector potential of houseflies (Musca domestica)forHelicobacter pylori. J. Clin. Microbiol. 35, 1300–1303. 29. Shane, S. M., Montrose, M. S., and Harrington, K. S. (1985). Transmission of Campylobacter jejuni by the housefly (Musca domestica). Avian Dis. 29(2), 384–391. 30. Kobayashi, M., Sasaki, T., Saito, N., Tamura, K., Suzuki, K., Watanabe, H., and Agui, N. (1999). Houseflies: not simple mechanical vectors of enterohemorrhagic Escherichia coli O157: H7. Am. J. Trop. Med. Hyg. 61(4), 625–629. 31. Moriya, K., Fujibayashi, T., Yoshihara, T., Matsuda, A., Sumi, N., Umezaki, N., Kurahashi, H., Agui, N., Wada, A., and Watanabe, H. (1999). Verotoxin-producing Escherichia coli O157: H7 carried by the housefly in Japan. Med. Vet. Entomol. 13(2), 214–216. 32. Sasaki, T., Kobayashi, M., and Agui, N. (2000). Epidemiological potential of excretion and regurgitation by Musca domestica (Diptera: Muscidae) in the dissemination of Escherichia coli O157: H7 to food. J. Med. Entomol. 37(6), 945–949. 33. Mian, L. S., Maag, H., and Tacal, J. V. (2002). Isolation of Salmonella from muscoid flies at commercial animal establishments in San Bernardino County, California. J. Vector Ecol. 27(1), 82–85. 34. Nayduch, D., Noblet, G. P., and Stutzenberger, F. J. (2002). Vector potential of houseflies for the bacterium Aeromonas caviae. Med. Vet. Entomol. 16(2), 193–198. 35. Ahmad, A., Nagaraja, T. G., and Zurek, L. (2007). Transmission of Escherichia coli O157: H7 to cattle by house flies. Prev. Vet. Med. 80(1), 74–81. 36. Janisiewicz, W. J., Conway, W. S., Brown, M. W., Sapers, G. M., Fratamico, P., and Buchanan, R. L. (1999). Fate of Escherichia coli 0157: H7 on fresh-cut apple tissue and its potential for transmission by fruit flies. Appl. Environ. Microbiol. 65(1), 1–5. 37. Sela, S., Nestel, D., Pinto, R., Nemny-Lavy, E., and Bar-Joseph, M. (2005). Mediter- ranean fruit fly as a potential vector of bacterial pathogens. Appl. Environ. Microbiol. 71(7), 4052–4056. 38. Cohen, D., Green, M., Block, C., Slepon, R., Ambar, R., Wasserman, S. S., and Levine, M. M. (1991). Reduction of transmission of shigellosis by control of houseflies (Musca domestica). Lancet 337(8748), 993–997. 39. Esrey, S. A. (1991). Interventions for the Control of Diarrhoeal Diseases Among Young Children: Fly Control, World Health Organization, Geneva, Published document WHO/CDD/91.37. 40. Macovei, L., and Zurek, L. (2006). Ecology of antibiotic resistance genes: characterization of enterococci from houseflies collected in food settings. Appl. Environ. Microbiol. 72(6), 4028–4035. 41. Macovei, L., and Zurek, L. (2007). Influx of enterococci and associated antibiotic resistance and virulence genes from ready-to-eat food to the human digestive tract. Appl. Environ. Microbiol. 73(21), 6740–6747. 42. Gurney, A. B., Fisk, F. W. (1991). Cockroaches. In Agriculture Handbook 655 , Insect and Mite Pests in Food: An Illustrated Key, J. R. Gorham, Ed. Superintendent of Documents, U. S. Government Printing Office, Washington, DC, pp. 45–74, 527–544. Voeller V05-c07.tex V1 - 12/04/2013 1:06pm Page 107

REFERENCES 107

43. Robinson, W. H. (2005). Urban Insects and Arachnids, Cambridge University Press, Cam- bridge. 44. Bracke, J. W., Cruden, D. L., and Markovetz, A. J. (1979). Intestinal microbial flora of the American cockroach, periplaneta American L. Appl. Environ. Microbiol. 38(5), 945–955. 45. Cruden, D. L., and Markovetz, A. J. (1987). Microbial ecology of the cockroach gut. Annu. Rev. Microbiol. 41, 617–643. 46. Roth, L. M., and Willis, E. R. (1960). The biotic associations of cockroaches. Smithson Misc. Coll. 141, 1–470. 47. Bennett, G. (1993). Cockroaches as carriers of bacteria. Lancet 341(8847), 732. 48. Cotton, M. F., Wasserman, E., Pieper, C. H., Theron, D. C., van Tubbergh, D., Camp- bell, G., Fang, F. C., and Barnes, J. (2000). Invasive disease due to extended spectrum beta-lactamase-producing Klebsiella pneumoniae in a neonatal unit: the possible role of cockroaches. J. Hosp. Infect. 44(1), 13–17. 49. Graffar, M., and Mertens, S. (1950). Le roleˆ des blattes dans la transmission des salmonelloses. Ann. Inst. Pasteur 79, 654–660. 50. Gazivoda, P., and Fish, D. (1985). Scanning electron microscope demonstration of bacteria on the tarsi of Blattella germanica. J. N. Y. Entomol. Soc. 93, 1064–1067. 51. Roth, L. M., and Willis, E. R. (1957). The medical importance of cockroaches. Smithson Misc. Coll. 134(10), 1–147. 52. Elgderi, R. M., Ghenghesh, K. S., and Berbash, N. (2006). Carriage by the German cockroach (Blattella germanica) of multiple-antibiotic-resistant bacteria that are potentially pathogenic to humans, in hospitals and households in Tripoli, Libya. Ann. Trop. Med. Parasitol. 100(1), 55–62. 53. Tatfeng, Y. M., Usuanlele, M. U., Orukpe, A., Digban, A. K., Okodua, M., Oviasogie, F., and Turay, A. A. (2005). Mechanical transmission of pathogenic organisms: the role of cockroaches. J. Vector Borne Dis. 42(4), 129–134. 54. Mpuchane, S., Allotey, J., Matsheka, I., Simpanya, M., Coetzee, S., Jordaan, A., Mrema, N., and Gashe, B. A. (2006). Carriage of micro-organisms by domestic cockroaches and implications for food safety. Int. J. Trop. Insect. Sci. 26, 166–175. 55. Chaichanawongsaroj, N., Vanichayatanarak, K., Pipatkullachat, T., Poirojpanya, M., and Somkiatcharoen, S. (2004). Isolation of gram-negative bacteria from cockroaches trapped from urban environment. Southeast Asian J. Trop. Med. Public Health 35(3), 681–684. 56. Prado, M. A., Gir, E., Pereira, M. S., Reis, C., and Pimenta, F. C. (2006). Profile of antimicrobial resistance of bacteria isolated from cockroaches (Periplaneta Americana) in a Brazilian health care institution. Braz.J.Infect.Dis.10(1), 26–32. 57. Pai, H.-H., Chen, W. C., and Peng, C. F. (2004). Cockroaches as potential vectors of nosocomial infections. Infect. Control Hosp. Epidemiol. 25(11), 979–984. 58. Pai, H.-H., Chen, W. C., and Peng, C. F. (2003). Isolation of non-tuberculous mycobacteria from hospital cockroaches (Periplaneta Americana). J. Hosp. Infect. 53, 224–228. 59. Lemos, A. A., Lemos, M. A., Prado, M. A., Pimenta, F. C., Gir, E., Silva, H. M., and Silva, M. R. R. (2006). Cockroaches as carriers of fungi of medical importance. Mycoses 49(1), 23–25. 60. Pai, H.-H., Ko, Y. C., and Chen, E. R. (2003). Cockroaches (Periplaneta Americana and Blattella germanica) as potential mechanical disseminators of Entamoeba histolytica. Acta Trop. 87(3), 355–359. 61. Smith, D. R. (1991). Ants (Formicidae, Hymenoptera). In Agriculture Handbook 655 , Insect and Mite Pests in Food: An Illustrated Key, J. R. Gorham, Ed. Superintendent of Documents, U. S. Government Printing Office, Washington, DC, pp. 297–309, 633–649. Voeller V05-c07.tex V1 - 12/04/2013 1:06pm Page 108

108 INSECTS AS VECTORS OF FOODBORNE PATHOGENS

62. Smith, M. R. (1965). House-infesting Ants of Eastern United States, Technical Bulletin 1326, U. S. Department of Agriculture, Washington, DC. 63. Beatson, S. (1973). Pharaoh’s ants enter giving sets. Lancet 1(7803), 606. 64. Cartwright, R. Y., and Clifford, C. M. (1973). Pharaoh’s ants. Lancet 2(7843), 1455–1456. 65. Boursaux-Eude, C., and Gross, R. (2000). New insights into symbiotic associations between ants and bacteria. Res. Microbiol. 151(7), 513–519. 66. Zientz, E., Feldhaar, H., Stoll, S., and Gross, R. (2005). Insights into the microbial world associated with ants. Arch. Microbiol. 184, 199–206. 67. Hughes, D. E., Kassim, O. O., Gregory, J., Stupart, M., Austin, I., and Duffield, R. (1989). Spectrum of bacterial pathogens transmitted by Pharaoh’s ants. Lab. Anim. Sci. 39(2), 167–168. 68. Ipinza-Regla, J., Figueroa, G., and Moreno, I. (1984). Iridomyrmex humilis (Formicidae) y su papel como possible vector de contaminacion´ microbiana en industrias de alimentos. Folia Entomol. Mex. 62, 111–124. 69. de Zarzuela, M. F. M., Campos-Farinha, A. E. C., and Peçanha, M. P. (2005). Evaluation of urban ants (Hymenoptera: Formicidae) as carriers of pathogens in residential and industrial environments. Sociobiology 45(1), 9–14. 70. Beatson, S. H. (1972). Pharaoh’s ants as pathogen vectors in hospitals. Lancet 1(7747), 425–427. 71. Chadee, D. D., and Le Maitre, A. (1990). Ants: potential mechanical vectors of hospital infections in Trinidad. Trans. R. Soc. Trop. Med. Hyg. 84, 297. 72. da Costa, S. B., Pelli, A., de Carvalho, G. P., Oliveira, A. G., da Silva, P. R., Teixeira, M. M., Martins, E., Terra, A. P. S., Resende, E. M., Hueb, C. C., de Oliveira, B., and de Morais, C. A. (2006). Ants as mechanical vectors of microorganisms in the school hospital of the universidade federal do Trianguloˆ Mineiro. Rev. Soc. Bras. Med. Trop. 39(6), 527–529. 73. Edwards, J. P., and Baker, L. F. (1981). Distribution and importance of the Pharaoh’s ant Monomorium pharaonis (L.) in National Health Service Hospitals in England. J. Hosp. Infect. 2(3), 249–254. 74. Fowler, H. G., Bueno, O. C., Sadatsune, T., and Montelli, A. C. (1993). Ants as potential vectors of pathogens in hospitals in the State of Sao˜ Paulo, Brazil. Insect Sci. Appl. 14, 367–370. 75. Harein, P. K., and De Las Casas, E. (1968). Bacteria from granary weevils collected from laboratory colonies and field infestations. J. Econ. Entomol. 61(6), 1719–1720. 76. Dunkel, F. V. (1988). The relationship of insects to the deterioration of stored grain by fungi. Int. J. Food Microbiol. 7, 227–244. 77. Husted, S. R., Mills, R. B., Foltz, V. D., and Crumrine, M. H. (1969). Transmission of Salmonella montevideo from contaminated to clean wheat by the rice weevil. J. Econ. Ento- mol. 62(6), 1489–1491. 78. Eugenio, C., De Las Casas, E., Harein, P. K., and Mirocha, C. J. (1970). Detection of the mycotoxin F-2 in the confused flour beetle and the lesser mealworm. J. Econ. Entomol. 63(2), 412–415. 79. Pande, N., and Mehrotra, B. S. (1988). Rice weevil (Sitophilus oryzae Linn.): vector of toxigenic fungi. Nat. Acad. Sci. Lett. (India) 11, 3–4.

FURTHER READING

Agbodaze, D., and Owusu, S. B. (1989). Cockroaches (Periplaneta Americana)ascarriersof agents of bacterial diarrhoea in Accra, Ghana. Cent.Afr.J.Med.35(9), 484–486. Voeller V05-c07.tex V1 - 12/04/2013 1:06pm Page 109

FURTHER READING

Iowa State University The Center for Food Security and Public Health website, http://www. cfsph.iastate.edu/ National Research Council of the National Academies (2005). Animal Health at the Crossroads: Preventing, detecting and diagnosing animal diseases. The National Academies Press, Wash- ington, DC. Voeller V05-c09.tex V1 - 12/04/2013 1:07pm Page 133

9 POTENTIAL FOR HUMAN ILLNESS FROM ANIMAL TRANSMISSION OF FOOD-BORNE PATHOGENS

David M. Hartley Department of Radiology, Georgetown University School of Medicine, Washington, DC

9.1 INTRODUCTION

Humans become infected with pathogens via aerosol, oral, and percutaneous pathways. Infection may produce a spectrum of outcomes, ranging from asymptomatic and self-limited disease, to long-term sequelae, to death. During the Cold War, several nations exploited these facts to develop biological weapons for use against humans. Designed to achieve specific strategic or tactical military objectives efficiently, weapons were based on a limited number of agents and designed to be delivered primarily as small particle aerosols [1, 2]. In contrast, bioterrorists may employ a multiplicity of microbial agents, delivered via diverse pathways, against civilian populations. The diversity of entryway and a spectrum of outcomes makes bioterrorism a difficult problem to characterize and defend against. In this chapter we focus on diseases of humans associated with animals or food and foodstuffs, which have demonstrated potential to disrupt populations and societies. Human illness could result from a bioterrorist infecting animal species (e.g. livestock, wildlife, and insect vectors) or food and foodstuffs with biological agents aimed at human populations, or as “collateral damage” in a biological attack aimed at domestic livestock. Volumes have been written on zoonotic and food-borne illnesses; this is a brief resume. However broad, three underlying themes are evident: (i) the threat spectrum is very broad; (ii) options for disease control are diverse and often problematic; and (iii) there is a need for creative approaches to both threat analysis and control and prevention.

9.2 SCIENTIFIC OVERVIEW

This is a brief overview of zoonotic and food-borne threats. Details regarding individual agents and diseases can be found in the References and Further Reading sections.

133 Voeller V05-c09.tex V1 - 12/04/2013 1:07pm Page 134

134 POTENTIAL FOR HUMAN ILLNESS FROM ANIMAL TRANSMISSION

9.2.1 Zoonotic Threats 9.2.1.1 Scope of the Problem. Diseases of humans acquired from animal sources or reservoirs—zoonoses—have figured prominently in human health historically. For example, Yersinia pestis, the causative agent of plague, is carried by different rodents, lagomorphs, and feline species and vectored to humans by fleas. In the fourteenth century an epidemic of bubonic and pneumonic plague caused the death of an estimated one-third of the population of Europe [3]. Reverberations from this event lasted for centuries; even today the word plague carries the connotation of a disastrous affliction.

9.2.1.2 Agents. The United States Department of Health and Human Services (DHHS) has published a list of agents important for biodefense (Table 9.1), many of which are zoonotic. The United States Department of Agriculture (USDA) has an analogous list of agents relevant to livestock health, the majority of which do not cause human disease. The intersection of these describes some 20 zoonotic pathogens of public health importance. The threat to human health, however, is much broader. A standard reference describes approximately 150 zoonoses and communicable diseases common to man and animals, and new and novel zoonoses continue to be recognized [4–8]. Table 9.2 contains additional threat agents that do not appear on the HHS and USDA lists, but nonetheless command respect.

9.2.1.3 Transmission. Many zoonotic agents infect humans via multiple pathways. Y. pestis, for example, can be transmitted from rodents to humans via biting ﬂeas, droplets, and aerosols. Rift Valley fever (RVF) virus is transmitted to humans from ungulate species via mosquitoes, via aerosolized blood or body ﬂuids from a viremic animal or abortus, and possibly through contact with infected meat. Bacillus anthracis, which often spreads from animals to humans via direct contact, can also infect humans via contaminated food and inhalation. Francisella tularensis can infect humans via biting arthropods, aerosols, or by handling and consuming infected meat and water. Severity of illness can depend upon the route of infection. In the case of B. anthracis, for example, cutaneous anthrax has a much lower case fatality rate (CFR; ∼20%) than inhalation anthrax (∼90%) [10]. Similar remarks apply to bubonic versus pneumonic plague and vector-borne versus pneumonic tularemia.

9.2.1.4 Disease Ecology. Many zoonoses are transmitted in distinct areas suited to their transmission and persistence. If such agents are translocated to new geographic areas, they may be successful if ecologic conditions are suitable. Many factors contribute to suitability, including inter alia land use, host population abundance, climate, presence and density of reservoir species, and vector capacity among indigenous arthropods. For example, in the 1999 North American introduction of West Nile virus (WNV), the agent found a diverse set of wildlife hosts and competent mosquito vectors, facilitating effective local and long-range spread (and thereby increasing human morbidity). Other examples of the translocation of pathogens to immunologically na¨ıve regions include Y . pestis, which entered the Western United States in the late nineteenth century, finding wildlife and domestic hosts, and persists to the present; and RVF virus, which entered the Arabian Peninsula in 2000, and continues to circulate [11]. In general, the interactions between, and the relative importance and significance of, specific ecologic factors will differ from agent to agent. Voeller V05-c09.tex V1 - 12/04/2013 1:07pm Page 135

SCIENTIFIC OVERVIEW 135

TABLE 9.1 HHS Select Agents and Toxins

Overlap Agents and Toxins (common Non-Overlap Agents and Toxins (HHS list only) to HHS and USDA lists)

Bacteria, Fungi, and Rickettsia Coccidioides posadasii Bacillus anthracisa Rickettsia prowazekii a Botulinum neurotoxin producing species of Clostridium Rickettsia rickettsii a Brucella abortusa Yersinia pestisa Brucella melitensisa Brucella suisa Burkholderia mallei (formerly Pseudomonas mallei)a Burkholderia pseudomallei (formerly Pseudomonas pseudomallei)a Coccidioides immitisa Coxiella burnetii a Francisella tularensisa

Toxins Abrin toxin Botulinum neurotoxins Conotoxins Clostridium perfringens epsilon toxin Diacetoxyscirpenol Shigatoxin Ricin Staphylococcal enterotoxins Saxitoxin T-2 toxin Shiga-like ribosome inactivating proteins Tetrodotoxin

Viruses Cercopithecine herpesvirus (Herpes B virus) Eastern Equine Encephalitis virusa Crimean-Congo hemorrhagic fever virusa Hendra virusa Ebola Virusa Nipah Virusa Lassa fever virusa Rift Valley fever virusa Marburg virusa Venezuelan Equine Encephalitis virusa Monkeypox virusa Reconstructed replication competent forms of the 1918 pandemic influenza virus containing any portion of the coding regions of all eight gene segments (Reconstructed 1918 Influenza virus) South American hemorrhagic fever viruses (Flexal, Guanarito, Junin, Machupo, Sabia)a Tick-borne encephalitis complex (flavi) viruses (Central European tick-borne encephalitis, Far Eastern tick-borne encephalitis, Kyasanur forest disease, Omsk hemorrhagic fever, Russian spring and summer encephalitis)a Variola major virus (Smallpox virus) Variola minor virus (Alastrim) [9] aZoonotic agents Voeller V05-c09.tex V1 - 12/04/2013 1:07pm Page 136

136 POTENTIAL FOR HUMAN ILLNESS FROM ANIMAL TRANSMISSION

TABLE 9.2 Selected Additional Zoonotic Agents Posing Threats to Human Health

Borrelia burgdorferi Campylobacter jejuni Leptospira interrogans Staphylococcus aureus

Viruses Rabies Mosquito-borne ﬂaviviruses (West Nile, Japanese Encephalitis) Sin nombre Yellow fever Emerging and yet-to-emerge inﬂuenza SARS coronavirus

Prion vCJD prion

Parasite Trypanosoma cruzi

A zoonosis introduced into an immunologically na¨ıve population may behave dif- ferently, epidemiologically, than it does in endemic regions [12]. In the case of WNV in North America, for example, disease in both humans and wildlife species was more severe than recent observations in Europe and the Middle East would have suggested [13]. Transmission and disease severity are complex functions of environment, host resistance, and immunity, and properties of the agents themselves (all of which are dynamic).

9.2.1.5 Disease Diversity. As a whole, zoonotic agents cause a spectrum of disease. Toward the less virulent end of the spectrum is WNV, in which the proportion of asymptomatic infections may be as high as 80%; neuroinvasive disease occurs in a small minority of cases. More virulent RVF is rarely asymptomatic, typically resulting in self-limited febrile illness; a minority of cases results in severe complications including hepatitis, retinal hemorrhage, and hemorrhagic fever. At the most virulent end of the spectrum are Marburg virus and Ebola virus, fatal in as much as 90% of all cases.

9.2.1.6 Therapy and Prevention. Antimicrobial and vaccine prophylaxis is available for a subset of the agents shown in Tables 9.1, 9.2. Antibiotic resistance is observed in some pathogens. Multidrug-resistant Y. pestis has been observed in outbreaks in Mada- gascar [17]. Vaccines exist for a small subset of the zoonoses (Table 9.3); many possess poor epidemiologic properties (e.g. low protective efﬁcacy, short period of protection) and undesirable side effects (e.g. teratogenisis, severe local tissue reaction). Personal protective measures including barrier precautions, breathing apparatus, and vector repellent can have good protective efﬁcacy; which are appropriate are agent- and scenario-dependent. At the population level, public health measures including vector and reservoir species control can be effective. However, such measures can have unintended consequences and should be considered with respect to the scenario. Theoretically, for example, indiscrim- inant rodent culling in plague-endemic areas could result in a shortage of natural hosts Voeller V05-c09.tex V1 - 12/04/2013 1:07pm Page 137

SCIENTIFIC OVERVIEW 137

TABLE 9.3 Human Vaccines for Zoonotic Threat Agents

Agent Vaccine Type Status

Eastern equine encephalitis Inactivated INDa, but no longer produced virus Japanese Encephalitis virus Inactivated US licensed vaccine Junin virus Live attenuated Not US licensed. Meets requirements for IND application Kyasanur virus Inactivated Not US licensed. Shown to be effective in ﬁeld trials in India Monkeypox virus Live attenuated Same as smallpox virus vaccine (and thus part of the US Strategic National Stockpile) Rabies virus Inactivated US licensed vaccine Rift Valley fever virus Inactivated IND Live attenuated Under IND Tick-borne encephalitis virus Inactivated Not US licensed. Available in Europe Yellow fever virus Live attenuated US licensed Venezuelan equine Live attenuated IND, but no longer produced encephalitis virus Inactivated IND, but no longer produced Bacillus anthracis Inactivated US licensed vaccine. Part of the US Strategic cell-free National Stockpile Coxiella burnetii Inactivated IND, but no longer produced Francisella tularensis Live attenuated Not currently available in the US Yersinia pestis Inactivated Effective against bubonic but not pulmonary plague. No longer produced aIND is required in the United States before a product undergoes human testing. [See 22 - Refs 3, 10, 14–16, Chapter 3] for vectors carrying Y. pestis [18]. Fleas abandoning dead rodents could instead infest human habitations, spawning transmission of bubonic plague. In other cases it may be possible to protect humans by controlling disease in intermediate hosts. For example, in the case of RVF it is theorized that the most effective approach for human protection is to vaccinate domestic livestock populations.

9.2.2 Food-Borne Threats 9.2.2.1 Scope of the Problem. Modern food production and storage technology has caused massive declines in food-borne illness. Nonetheless, outbreaks continue to occur, facilitated by, inter alia, centralized food production and supply systems; the growing frequency of imported food and foodstuffs from nations with less-developed agricultural and production procedures and practices; and lapses in good preparation and serving practices in dining establishments and homes. These and related avenues represent potential targets of bioterrorism [19].

9.2.2.2 Agents. Foods ready to eat as well as their ingredients can carry a large number of pathogens and toxins. There is no analog of the DHHS select agent list for food-borne pathogens; Table 9.4 contains a representative collection. Generally, pathogens must Voeller V05-c09.tex V1 - 12/04/2013 1:07pm Page 138

138 POTENTIAL FOR HUMAN ILLNESS FROM ANIMAL TRANSMISSION pass the gastric barrier successfully and colonize the gut to produce disease. Disease can (but need not) be toxin-mediated, meaning that toxins are produced as a by-product of microbial growth in the gut. Foods contaminated with toxins at the time of consumption can produce disease regardless of microbial growth in the gut (food-borne intoxication).

9.2.2.3 Transmission. Poor food sanitation and handling practices are important causes of disease. Secondary transmission via fecal—oral routes is possible for many viral and bacterial pathogens, due to a combination of high levels of pathogen shedding, viability and persistence of the agents in the environment and the low infectious dose. For example, depending on the age and condition of host, the infective dose for Shigella species may be as low as 10–100 cells. The ID50 (dose sufﬁcient to infect 50% of exposed persons) infectious dose of Vibrio cholerae maybeashighas103 organisms, but recent observations imply a “hyperinfectious” state of the organism in which this is reduced by 1–2 orders of magnitude for a short time after the pathogen is shed by the host [20]. Whether hyperinfectious states exist for other enteric pathogens of humans remains unknown. Generally, the ID50 of food-borne organisms are incompletely known and variable, depending, inter alia, upon the food matrix within which they are consumed and the acidity of the gastric environment. Toxins can be extremely potent; botulinum toxin A (a protein neurotoxin), for example, possesses a mean lethal dose <0.001 μg/kg and has been called the most lethal substance known to humans [21].

9.2.2.4 Disease Ecology. Many enteric pathogens are ubiquitous in the natural environment. Clostridium botulinum, for example, is found commonly in soil and can contaminate food or foodstuffs anytime between harvest and preparation. C. botulinum spores can germinate and grow, producing toxins, in anaerobic environments. Home and occasionally commercially canned products have been implicated in botulism cases; other foods associated with botulism include herb-infused oils, bottled garlic, and baked potatoes in aluminum foil that have been held warm for extended periods of time. C. botulinum subtypes may produce one or more of seven types of toxin that readily cross the gastric acidic barrier. Vibrio species (including V . cholerae, Vibrio parahaemolyticus and Vibrio vulniﬁcus) are common in estuarine and marine environments, frequently contaminating water and shellﬁsh. V. cholerae is also a major cause of disease in areas lacking effective water treatment and sanitation systems.

9.2.2.5 Disease Diversity. Food-borne agents cause a broad spectrum of disease. Toward the less virulent end of the spectrum are the Norwalk-like viruses and enterotoxigenic (ETEC) strains of Escherichia coli (CFRs in nonpediatric and nongeriatric populations near 1%). More virulent are the enterohemorrhagic Escherichia coli (EHEC) and other shigatoxin-producing Escherichia coli (STEC), often associated with hemolytic uremic syndrome (HUS), kidney failure, and death. Toward the most virulent end of the spectrum are Shigella dysenteriae and E. coli O157:H7, associated HUS; Campylobacter jejuni, associated with Guillain—Barre syndrome; and variant Creutzfeldt—Jakob disease (vCJD) prion (PrPSc), which may be latent for years but conveys near 100% mortality.

9.2.2.6 Therapy and Prevention. Antibiotics may be indicated for some bacterial infections (e.g. Listeria monocytogenes, Brucella spp., B. anthracis) but are contraindicated for many others (e.g. EHEC). Many agents are developing antibiotic resistance. For viral Voeller V05-c09.tex V1 - 12/04/2013 1:07pm Page 139

SCIENTIFIC OVERVIEW 139

TABLE 9.4 Selected Food- and Water-borne Agents Posing Threats to Human Health Bacteria Bacillus anthracis, B. cereus Brucella species Campylobacter jejuni Clostridium botulinum, C. perfringens Enterovirulent Escherichia coli (including EHEC, ETEC, EPEC, and EIEC strains) Listeria monocytogenes Salmonella typhi Shigella species Staphylococcus aureus Streptococcus pyogens Vibrio cholerae, V. parahaemolyticus, V. vulniﬁcus Yersinia enterocolitica

Viruses Calciviruses Hepatitis A virus Rotovirus Astroviruses Adenoviruses Parvoviruses

Parasites Cryptosporidium parvum Cyclospora cayetanensis Giardia lamblia Toxoplasma gondii

Toxins Anatoxin A Botulinum toxin Ciguatera toxin Epsilon toxin of C. perfringens Staphylococcal enterotoxins Saxitoxin T-2 mycotoxin Tetrodotoxin Ref. [21]

diseases and nonviral diseases if antibiotics are contraindicated, therapy is limited to supportive measures. Vaccines exist for very few agents as seen in Table 9.4. A rotovirus vaccine was approved for infants in the United States in 1998 but was removed in 1999 when it was found to be associated with intussusception. Currently, available cholera vaccine provides weak (∼50%) protection 3–6 months after administration; new vaccines are under development [22]. Investigational vaccines exist for ETEC and Shigella strains. Investigational antitoxins exist for botulinum toxin, but must be given early in the course of the disease. Voeller V05-c09.tex V1 - 12/04/2013 1:07pm Page 140

140 POTENTIAL FOR HUMAN ILLNESS FROM ANIMAL TRANSMISSION

Given the limited chemoprophylactic and chemotherapeutic options, preventive measures are central to avoiding food-borne illnesses. Food inspection and good practice in food sanitation and hygiene may be the ﬁrst line of defense in preventing an introduced food-borne illness, especially secondary cases of infection. With few exceptions, proper food manufacturing, storage, and preparation practices can prevent illness. Exceptions include Staphylococcus aureus and some Bacillus species, which produce heat-stable toxins. At the population level, the “hazard analysis and critical control point” (HACCP) approach has proven effective in reducing the microbial load of foods and foodstuffs, and thus the incidence of disease. As the comments of this section suggest, a bioterrorist attack on livestock could result in human illness via multiple pathways, including vector-borne transmission (e.g. RVF, eastern equine encephalitic (EEE)), aerosols (e.g. RVF), infected meat or milk (e.g. brucellosis), and direct contact (e.g. anthrax, glanders). People at primary risk would include agricultural and abattoir workers and those associated with handling animals and raw foodstuffs. Introduction of multiple agents into wildlife species could similarly result in human disease in anyone exposed to wildlife or wildlife-feeding vectors. Regarding food-borne attacks, proper production, preservation, and preparation practices can safeguard against contamination (but may not render heat-stable toxins inert). Post-production contamination, or contamination of foods that are traditionally not cooked or are eaten raw (e.g. lunch meat, fruits, and vegetables), may have a high probability of producing human disease.

9.3 CURRENT RESEARCH

In this section we highlight research directly related to detection, control, and prevention of human illness events (i.e. sporadic cases or outbreaks).

9.3.1 Detection Early outbreak detection can reduce human illness. There are two essential approaches to event detection: detection by health care providers and detection via the exploitation of surveillance data. In the first category, the incidence of disease in individual health care facilities is detected by clinical laboratory tests. For most organisms of interest, sensitive, and specific microbiologic tests provide results within minutes to days of specimen collection. Higher performance tests are under development (discussed elsewhere in this Handbook). Typically, infections are noted by periodic manual chart review or computer data mining and reported, as appropriate, to public health agencies. In the United States, “notifiable diseases” vary between states; if a particular disease is not notifiable, cases may be reported weeks or months later, and possibly not at all. Surveillance is the “ongoing, systematic collection, analysis, and interpretation of outcome-specific health data, coupled with the timely dissemination of these data and their analyses” to medical, public health, and homeland security stakeholders [23]. Surveillance can entail the active or passive collection of data such as laboratory-confirmed cases (the most common), incidence of disease syndromes, human behavior thought to be associated with outbreaks, or environmental specimens. In the United States, public health surveillance data regarding many of the agents discussed in this chapter are reported to different Centers for Disease Control and Prevention (CDC) national surveillance Voeller V05-c09.tex V1 - 12/04/2013 1:07pm Page 141

CURRENT RESEARCH 141 systems (e.g. the National Notifiable Diseases Surveillance System [NNDSS] and the Food-borne Disease Active Surveillance Network [FoodNet]). No single surveillance system in the United States conducts surveillance for all bioterrorism-related agents or conditions [24]. A recent study, for example, identified more than 20 US-based networks for food-borne diseases alone [25]. Syndromic surveillance and human behavior/social disruption surveillance are under active development at present [26, 27]. Surveillance of air samples is carried out in various US cities. Specimens from col- lectors are removed periodically and analyzed for the presence of certain agents [28]. Details regarding this system, known as BioWatch, are scarce. However, the system has produced a number of alerts in different cities, all focused on the zoonotic agent F. tularensis. Domestic or wild animals can serve as sentinels of infection (e.g. avian influenza virus [29] and arboviruses [30]). A recent analysis [31] reviewed the potential of animals as sentinels of bioterrorism agents, concluding that “for certain bioterrorism agents, pets, wildlife, or livestock could provide early warning” of an attack. Event detection from surveillance data is a complex function of measured observables and analytic methodology. The main objective is to identify a signal corresponding to a biological event of interest in a noisy and often ill-defined background. Mathematical and statistical approaches to analyzing such data at present include: spatial, temporal, and spatio—temporal clustering analyses; data mining; multivariable monitoring; and data fusion. Some approaches can be automated and/or executed in real- or near real-time. A review the statistical issues and challenges associated with timely integration of multiple data sources for purposes of detecting epidemiologic events can be found in reference [32, 33]

9.3.2 Control and Intervention Control and intervention research focuses on discovering new chemoprophylactic and chemotherapeutic agents and identifying public health measures. Gaps revealed in Table 9.3 are the reason for current emphasis on developing new and improved vaccines. Such efforts, however, are not as vigorous as they could be due to the economic situation facing pharmaceutical companies. Lang and Wood [34] note that “costs from research to licensure, the risks inherent in vaccine development (e.g. technological constraints, regulatory approval), and the short- and long-term market financial evaluations (e.g. net present value, return on investment) are key factors in the decision to develop a vaccine” for rare diseases. Despite efforts to offset these risks (e.g. the Orphan Drug Act of 1982, the BioShield Act of 2004), discovery and development of new vaccines in the United States remains depressed. Similar comments apply to antibiotics. To spur research and development in drug and other biodefense technology development, the US National Institutes of Health (NIH) and Department of Homeland Security (DHS) sponsor academic “Centers of Excellence” in bioterrorism defense research. Each center is a consortium focused on either a specific topic (DHS centers) or serving specific geographical regions (NIH centers). (See URLs listed in Further Reading.) Among others, vaccines for tularemia, RVF, and brucellosis are being studied at these centers. Additional researches focus upon diagnostic technology and analytic methodologies. Identifying effective (or finding new) public health measures involves a range of analytic activity. Traditional epidemiologic observational studies identifying infection risk factors are useful, particularly for newly recognized agents about which little is known. Mathematical modeling and simulation techniques have led to fundamental insights Voeller V05-c09.tex V1 - 12/04/2013 1:07pm Page 142

142 POTENTIAL FOR HUMAN ILLNESS FROM ANIMAL TRANSMISSION regarding contagion processes and the applied practice of public health [35, 36]. Mathe- matical models enable evaluation of control and intervention measures such as vaccination (e.g. impact of protective efficacy, time-to-protection, and rate of administration), vector control (e.g. impact of adulticides vs. larvicides, spraying frequency, spraying concentration), host culling (e.g. impact of the rate of culling, effects of incomplete culling), and reduction in exposure rates, without costly or difficult laboratory experiments. Models are built upon knowledge of the epidemiology of specific diseases. If properly constructed and analyzed, they can provide information on the effects of different con- trols or other quantities of interest. Thus, models can guide further laboratory research as well as provide defense planners with valuable guidance. At present, the literature reveals a paucity of applications to zoonoses. Current work by modelers at the DHS National Center of Excellence for Foreign Animal and Zoonotic Disease Defense [37] is noteworthy because it involves experts in human and agricultural medicine, human and agricultural epidemiology, microbiology, mathematics, ecology, geography, and information technology. The effort is addressing aspects of spatio—temporal vulnerability to RVF introduction, the relative payoff of possible control and intervention measures, and the economic impacts of an introduction into the United States. A related methodology is risk assessment (RA), which aims to identify what events can happen, their likelihood, and their consequences. RA grew out of methods developed for assessing health impacts from chemical exposure. Typically, there are four stages to RA: hazard identification, exposure assessment, dose-response assessment, and risk characterization. Qualitative RA is descriptive and indicates whether disease is likely under specified conditions of exposure, whereas quantitative RA provides numerical estimates of risk [38]. Both types are relevant for risk management and communication in biodefense. RA has been applied to several bioterrorism issues, including the risk of a RVF incursion and its persistence within the European Community [39]. However, methodological and data gaps exist that continue to limit researchers ability to complete comprehensive RAs for most biothreat agents. Regarding food safety and microbial contamination, microbial risk assessment (MRA) is under active development at present. MRA recognizes the biologic nature of organisms and accounts for it in the analysis. Typical facets of MRA include (i) the dynamic concentration of biologic agents due to growth and death; (ii) the heterogeneous distribution of biologic agents on/in food and water media; (iii) the potential for secondary (e.g. person-to-person) transmission; and (iv) the role played by immunity [40]. MRA studies have been carried out on a relatively small number of food-borne pathogens; the extent to which the methods developed for food-borne diseases can be utilized in assessing homeland security risk remains undemonstrated. There are substantial uncertainties regarding the use of animal models to inform human RAs; the viability and infectivity of pathogens in environmental media; and the use of potential sampling and detection methods for RA. To address these and related issues, the US DHS and the Environmental Protection Agency have recently founded the Center for Advancing MRA.

9.4 CRITICAL NEEDS

More is needed to assess the homeland security implications of the above discussion. Here, we consider three broad categories. Voeller V05-c09.tex V1 - 12/04/2013 1:07pm Page 143

CRITICAL NEEDS 143

9.4.1 Data Needs Robust, real-time surveillance systems capable of accurately and precisely estimating baseline values of human and veterinary disease incidence in both time and space are needed. In addition to looking “inward” at local, regional, and national disease activity, surveillance must look “outward” to identify the potential of threats being imported from distant regions. This requires up-to-date knowledge of disease transmission in foreign areas as well as data on the ecological conditions that might support the emergence and transmission of zoonoses. Analytic methods and effective concepts of operations to exploit early warnings of diseases effectively must be developed in conjunction with public health and homeland security stakeholders. Mathematical modeling and RA methodologies require a range of data on different threat agents, such as distributions of latency and incubation periods; identiﬁcation of the competent vectors of different agents in terms of temporal (e.g. seasonal) and spatial distributions; periods of disease, attack rates, and related epidemiological quantities, especially for exotic diseases; and dose-response functions. At present, such data are not often available. Both laboratory and ﬁeld studies are needed. In the case of livestock diseases, up to date data on the location or probable location of livestock populations are needed, as are reliable data on livestock transportation between farms, feedlots, stockyards, and slaughter facilities.

9.4.2 Analytical Needs While RA and mathematical modeling can both contribute to vulnerability analysis (e.g. identification of ecological areas, distribution nodes, or manufacturing systems that are particularly vulnerable to bioterrorism), only a small number of such analyses have appeared from the list of agents appearing in Tables 9.1–9.4. Analyses of the potential economic impact of these diseases must be elucidated. Only if such data are available that it will be possible to identify priorities and understand the need for vaccines and other public health prevention and control efforts. Comprehensive analyses involving more agents are needed; and it is vital that stakeholders and experts from all relevant technical fields participate. Collaboration from related fields (e.g. ecology, microbiology, and psychology) is needed to resolve challenges facing mathematical epidemiologic modeling, including identifying methods to integrate data collected at different temporal and spatial scales into models; identifying of surrogates for data that do not exist; refining the multimod- eling approach (i.e. integration of models into larger models); and finding how best to communicate the results of models to decision makers, policy makers, planners, and public health operations personnel. Similarly, approaches to model validation must be developed for the rare and foreign diseases of highest interest, for which limited data is available.

9.4.3 Basic Science Needs Questions regarding how a variety of zoonotic agents are maintained in natural populations; how they respond to changes in the environment; and what causes them to emerge and be transmitted remain unanswered. Lederberg and Shope [41] notes that “the precise ecological factors that lead to human infection . . . are murky, and textbook descrip- tions of the epidemiology of most zoonotic diseases are at best simplistic.” In order to Voeller V05-c09.tex V1 - 12/04/2013 1:07pm Page 144

144 POTENTIAL FOR HUMAN ILLNESS FROM ANIMAL TRANSMISSION effectively counter threats from these diseases, we must understand the ecology of such agents better.

9.5 RESEARCH DIRECTIONS

In the short term, a central question is how to bring current and nascent capabilities to bear on public health preparedness so that effective interventions can be made in the event of a bioterror incident. Research into public health systems engineering should have a high priority. For example, how might current sensors be employed (e.g. in animal and food settings) in optimal spatial and temporal sampling strategies to detect threats at the earliest time possible? And given detection, what are the appropriate actions to initiate, given existing treatment and prophylactic options and logistic limitations on delivery and administration? It is unclear whether, for example, the BioWatch system has been designed according to such concerns, or whether surveillance systems that are under development aim to address such questions. Such questions could be investigated with current analytic methods. A first step entails the refinement and validation of analytic tools. Mathematical approaches are, arguably, the only way to identify vulnerabilities, opportunities for prevention, and candidate control options in a systematic, unbiased way. However, models must be validated if they are to be credible. Validation is difficult for rare food-borne infections and zoonotic diseases that exist only in foreign nations. Prospective field studies in endemic areas must be undertaken in order to validate models of important threat agents. In the longer term, effective vaccines, antibiotics, and antivirals must become available if we are to increase public health preparedness. Research to identify generalizable programmatic approaches to commercialization of new solutions would be of immense use. Research from the US Centers of Excellence, for example, must be transitioned and commercialized if new products are to become available at the clinical level. One way forward may be to focus on public—private partnerships for purposes of technology transfer and product transition. However, the economic disincentives facing pharmaceutical companies must be solved if these products are ever to be mass produced.

ACKNOWLEDGMENTS

The author thanks Dr J. Glenn Morris and all the anonymous reviewers for suggestions strengthening this chapter.

REFERENCES

1. Patrick, W. (2001). Biological warfare scenarios. In Firepower in the Lab: Automation in the Fight Against Infectious Diseases and Bioterrorism,S.P.Layne,T.J.Beugelsdijk,andC.K. N. Patel, Eds. Joseph Henry Press, Washington, DC, pp. 215–223. 2. Alibek, K. (1999). Biohazard, Random House, New York, 3. Inglesby, T. V., Dennis, D. T., Henderson, D. A., Bartlett, J. G., Ascher, M. S., Eitzen, E., Fine, A. D., Friedlander, A. M., Hauer, J., Koerner, J. F., Layton, M., McDade, J., Osterholm, M. T., O’Toole, T., Parker, G., Perl, T. M., Russell, P. K., Schoch-Spana, M., and Tonat, K. (2000). Plague as a biological weapon: medical and public health management. Working Group on Civilian Biodefense. JAMA 283(17), 2281–2290. Voeller V05-c09.tex V1 - 12/04/2013 1:07pm Page 145

REFERENCES 145

4. Acha, P. N., and Szyfres, B. (2003). Zoonoses and Communicable Diseases Common to Man and Animals, Bacterioses and Mycoses, Vol. I , 3rd ed., Pan American Health Organization, Washington, DC. 5. Acha, P. N., and Szyfres, B. (2003). Zoonoses and Communicable Diseases Common to Man and Animals, Chlamydioses, Rickettsioses, and Viroses, Vol. II , 3rd ed., Pan American Health Organization, Washington, DC. 6. Acha, P. N., and Szyfres, B. (2003). Zoonoses and Communicable Diseases Common to Man and Animals, Parasitoses, Vol. III , 3rd ed., Pan American Health Organization, Washington, DC. 7. Lederberg, J., and Shope, R. E. (1992). Emerging Infections: Microbial threats to Health in the United States, National Academy Press, Washington, DC. 8. Smolinski, M. S., Hamburg, M. A., and Lederberg J., Eds. (2003). Microbial Threats to Health: Emergence, Detection, and Response, National Academies Press, Washington, DC. 9. US Code of Federal Regulations 7 CFR Part 331; 9 CFR Part 121; 42 CFR Part 73, http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm. 10. Inglesby, T. V., O’Toole, T., Henderson, D. A., Bartlett, J. G., Ascher, M. S., Eitzen, E., Friedlander, A. M., Gerberding, J., Hauer, J., Hughes, J., McDade, J., Osterholm, M. T., Parker, G., Perl, T. M., Russell, P. K., Tonat, K., Working Group on Civilian Biodefense. (2002). Anthrax as a biological weapon, 2002: updated recommendations for management. JAMA 287(17), 2236–2252. 11. CDC, (2000). Update: outbreak of rift valley fever - Saudi Arabia, August–November 2000. MMWR Morb. Mortal. Wkly. Rep. 49(43), 982–985. 12. Cartwright, F. F., and Biddiss, M. (2004). Disease and History, Gloucestershire, Phoenix Mill. 13. Kilpatrick, A. M., Kramer, L. D., Jones, M. J., Marra, P. P., and Daszak, P. (2006). West Nile virus epidemics in North America are driven by shifts in mosquito feeding behavior. PLoS Biol. 4(4), e82, Epub 2006 Feb 28. 14. Borio, L., Inglesby, T., Peters, C. J., Schmaljohn, A. L., Hughes, J. M., Jahrling, P. B., Ksiazek, T., Johnson, K. M., Meyerhoff, A., O’Toole, T., Ascher, M. S., Bartlett, J., Breman, J. G., Eitzen, E. M. Jr., Hamburg, M., Hauer, J., Henderson, D. A., Johnson, R. T., Kwik, G., Layton, M., Lillibridge, S., Nabel, G. J., Osterholm, M. T., Perl, T. M., Russell, P., Tonat, K., Working Group on Civilian Biodefense. (2002). Hemorrhagic fever viruses as biological weapons: medical and public health management. JAMA 287(18), 2391–2405. 15. Peters, C. J. (2002). The role of antivirals in responding to biological threats. In Knobler, S. L., Mahmoud, A. A. F. and Pray, L. A. Eds. Biological Threats and Terrorism: Assessing the Science and Response Capabilities, National Academy Press, Washington, DC. 16. Lemon, S., Thaul, S., Fisseha, S., and O’Maonaigh, H., Eds. (2002). Protecting Our Forces: Improving Vaccine Acquisition and Availability in the U.S. Military, National Academy Press, Washington, DC. 17. Chanteau, S., Ratsifasoamanana, L., Rasoamanana, B., Rahalison, L., Randriambelosoa, J., Roux, J., and Rabeson, D. (1998). Plague, a reemerging disease in Madagascar. Emerg. Infect. Dis. 4(1), 101–104. 18. Keeling, M. J., and Gilligan, C. A. (2000). Metapopulation dynamics of bubonic plague. Nature 407(6806), 903–906. 19. Wein, L. M., and Liu, Y. (2005). Analyzing a bioterror attack on the food supply: the case of botulinum toxin in milk. Proc. Natl. Acad. Sci. U.S.A. 102(28), 9984–9989. 20. Merrell, D. S., Butler, S. M., Qadri, F., Dolganov, N. A., Alam, A., Cohen, M. B., Calderwood, S. B., Schoolnik, G. K., and Camilli, A. (2002). Host-induced epidemic spread of the cholera bacterium. Nature 417(6889), 642–645. Voeller V05-c09.tex V1 - 12/04/2013 1:07pm Page 146

146 POTENTIAL FOR HUMAN ILLNESS FROM ANIMAL TRANSMISSION

21. Khan, A. S., Swerdlow, D. L., and Juranek, D. D. (2001). Precautions against biological and chemical terrorism directed at food and water supplies. Public Health Rep. 116(1), 3–14. 22. Dennehy, P. H. (2001). Active immunization in the United States: developments over the past decade. Clin. Microbiol. Rev. 14(4), 872–908. 23. Thacker, S. B., and Berkelman, R. L. (1992). History of public health surveillance. In Public Health Surveillance, W. Halperin, E. L. Baker, and R. R. Monson, Eds. Van Nostrand Reinhold, New York, pp. 1–15. 24. Chang, M., Glynn, M. K., Groseclose, S. L. (2003). Endemic, notifiable bioterrorism-related diseases United States, 1992–1999. Emerg. Infect. Dis. [serial online] 2003 May. Available from: URL: http://www.cdc.gov/ncidod/EID/vol9no5/02-0477.htm. Last accessed 6 November 2006. 25. Besser, J. M. (2006). Systems to detect microbial contamination of the food supply. In Address- ing Foodborne Threats to Health: Policies, Practices, and Global Coordination. IoM Board on Global Health, National Academy Press, Washington, DC, p. 179. 26. Yih, W. K., Caldwell, B., Harmon, R., Kleinman, K., Lazarus, R., Nelson, A., Nordin, J., Rehm, B., Richter, B., Ritzwoller, D., Sherwood, E., and Platt, R. (2004). Centers for disease control and prevention (CDC). National bioterrorism syndromic surveillance demonstration program. MMWR Morb. Mortal. Wkly. Rep. 53(Suppl.), 43–49. 27. Wilson, J. M., Polyak, M. G., Blake, J. W., Collmann, J. (2008). A heuristic indication and warning staging model for detection and assessment of biological events. JAMA. 15(2), 158–171. 28. Lim, D. V., Simpson, J. M., Kearns, E. A., and Kramer, M. F. (2005). Current and developing technologies for monitoring agents of bioterrorism and biowarfare. Clin. Microbiol. Rev. 18(4), 583–607. 29. Boltz, D. A., Douangngeun, B., Sinthasak, S., Phommachanh, P., Rolston, S., Chen, H., et al. (2005). H5N1 influenza viruses in Lao People’s democratic republic. Emerg. Infect. Dis. 12(10), 1593–1595. 30. Buckley, A., Dawson, A., and Gould, E. A. (2006). Detection of seroconversion to West Nile virus, Usutu virus and Sindbis virus in UK sentinel chickens. Virol. J. 3, 71. 31. Rabinowitz, P., Gordon, Z., Chudnov, D., Wilcox, M., Odofin, L., Liu, A., and Dein, J. (2006). Animals as sentinels of bioterrorism agents. Emerg. Infect. Dis. 12(4), 647–652. 32. Fienberg, S. E., and Shmueli, G. (2005). Statistical issues and challenges associated with rapid detection of bio-terrorist attacks. Stat. Med. 24(4), 513–529. 33. Bravata, D. M., McDonald, K. M., Smith, W. M., Rydzak, C., et al. (2004). Systematic review: surveillance systems for early detection of bioterrorism-related diseases. Ann. Intern. Med. 140(11), 910–922. 34. Lang, J., and Wood, S. C. (1999). Development of orphan vaccines: an industry perspective. Emerg. Infect. Dis. 5(6), 749–756. 35. Hethcote, H. W. (2000). The mathematics of infectious diseases. SIAM Rev. 42, 599–653. 36. McKenzie, F. E. (2000). Why model malaria? Parasitol. Today 16, 511–516. 37. Gaff, H. D., Hartley, D. M., and Leahy, N. P. (2007). An epidemiological model of Rift Valley fever virus. Elect. J. Diff. Equat. 115, 1–12. 38. Committee on Standards and Policies for Decontaminating Public Facilities Affected by Expo- sure to Harmful Biological Agents: How Clean is Safe? (2005). Reopening Public Facilities after a Biological Attack: A Decision-Making Framework, National Research Council, The National Academies Press, Washington, DC. 39. Pfeiffer, D., Pepin,´ M., Wooldridge, M., Schudel, A., Pensaert, M., Collins, D., Baldet, T., Davies, G., Kemp, A., Martin, V., Paweska, J., Swanepoel, R., and Thiongane, Y. (2005). The Voeller V05-c09.tex V1 - 12/04/2013 1:07pm Page 147

FURTHER READING

Additional information on these approaches to infectious disease surveillance can be found at the International Society for Disease Surveillance Web site, http://www.syndromic.org/, 2008. Projects under investigation at the DHS National Center of Excellence in Foreign Animal and Zoonotic Disease Defense (at Texas A&M University) are described at http://fazd.tamu.edu, 2008. Projects under investigation at the DHS National Center of Excellence for Food Protection and Defense (at the University of Minnesota) are described at http://www.ncfpd.umn.edu/, 2008. Information on the 10 National Institutes of Health Regional Centers of Excellence can be accessed via URL http://www3.niaid.nih.gov/research/resources/rce/, 2008. Projects under investigation at the DHS-EPA Center for Advancing Microbial Risk Assessment (at Michigan State University) are described at http://camra.msu.edu/, 2008. Banks, H. T., and Castillo-Chevez, C, Eds. (2003). Applications of a range of different mathematical modeling techniques are described. In Bioterrorism, SIAM, Philadelphia, PA. General information on the ecology and clinical symptoms of many of these diseases can be accessed at the WHO site http://www.who.int/mediacentre/factsheets/en/, 2008. Evans, A. S., and Brachman P. S., Eds. (1998). Detailed information on the ecology and clinical symptoms of many of these diseases appears. In Bacterial Infections of Humans, Plenum, New York. Evans, A. S., and Kaslow, R. A. (1997). Viral Infections of Humans, Plenum, New York, Though a bit dated, these are encyclopedic references. American Medical Association, American Nurses Association, American Nurses Foundation, Cen- ters for Disease Control and Prevention, Center for Food Safety and Applied Nutrition, U.S. Food and Drug Administration, Food Safety and Inspection Service, U.S. Department of Agri- culture. (2004). Diagnosis and management of foodborne illnesses: a primer for physicians and other health care professionals. MMWR Recomm. Rep. 53(RR-4), 1–33. Voeller V05-c09.tex V1 - 12/04/2013 1:07pm Page 148 Voeller V05-c10.tex V1 - 12/04/2013 1:07pm Page 149

10 MITIGATING CONSEQUENCES OF PATHOGEN INOCULATION INTO PROCESSED FOOD

James S. Dickson Department of Animal Science, Iowa State University, Ames, Iowa

10.1 INTRODUCTION

Pathogens occur in processed foods as a result of the natural occurrence of these organisms, and also as a result of failures in both processing and sanitation. The intentional introduction of pathogens into processed foods is an unlikely but plausible event, given an individual or individuals with sufficient motivation. The response to either an accidental or intentional event would be similar, although there would be more significance to an intentional event. However, the basic response would include the recovery of the affected products, a reevaluation of the process, and a public relations effort to restore public confidence in the specific food type or processor.

10.2 SCIENTIFIC OVERVIEW

10.2.1 Processed Foods The processing of foods ranges from very minimal to technologically advanced. An example of minimal processing would be fresh vegetables, such as green beans. In this case, the food is simply washed to remove physical contaminants, and may be prepackaged for retail sale. Alternatively, some foods, such as canned vegetables, may be processed and preserved to a degree that they are shelf stable and ready-to-eat. Because of the diversity of food types and processing, the potential for an intentional introduction of pathogens is great.

10.2.1.1 Historical Precedence. Upon leaving the post of Secretary of the US Depart- ment of Health and Human Services, Secretary Thompson commented “I, for the life of me, cannot understand why the terrorists have not, you know, attacked our food supply because it is so easy to do, and we are importing a lot of food from the Middle East, and it

149 Voeller V05-c10.tex V1 - 12/04/2013 1:07pm Page 150

150 MITIGATING CONSEQUENCES OF PATHOGEN INOCULATION INTO PROCESSED FOOD would be easy to tamper with that.” [1]. There have been historical examples of deliberate contamination of food products in the United States. Perhaps the largest incident was in State of Oregon in 1984 [2], where Salmonella was deliberately introduced into salad bars at local restaurants. In this case, the incident was intended to affect the outcome of a local election. In another incident, a disgruntled employee deliberately contaminated mufﬁns and doughnuts with Shigella dysenteriae and left them in an employee break room [3]. Potassium cyanide has been used to contaminate over-the-countermedications on several occasions, leading to the deaths of several unsuspecting consumers [4]. One of the earlier documented cases of this type of product tampering or contamination was an incident which occurred in Chicago in 1982, where seven people died after ingesting deliberately contaminated products [5]. Although some of these contamination events were attempts to gain ﬁnancially from the event (extortion), the motivation for the Chicago event is still unknown.

10.2.1.2 Potential Biological Agents. Biological agents that could be used as potential contaminants include bacteria, viruses, and parasites. In addition, microbial toxins could also be introduced into processed foods [6]. The information needed to select, isolate, identify, and cultivate these organisms is readily available within the public domain. Bacterial agents are more likely to be used as deliberate contamination agents, simply because a person with rudimentary microbiological skills and facilities, along with equipment and supplies purchased from readily available sources, can produce sufficient quantities of bacteria to be used to deliberately contaminate foods. The techniques and requirements to produce sufficient quantities of viruses and parasites would require more sophisticated skills and facilities. Bacterial agents may be divided into two general categories: those which cause infections and those which cause intoxications. Infectious agents require the presence of a live organism. Since most of these bacterial agents are sensitive to heat, they would be most effective with foods that do not undergo thermal processing or a final cooking step. The historical example of Salmonellae introduced into salad bars [2] illustrates this point. Bacterial toxins can also be produced, albeit in crude form, with rudimentary skills and facilities. Some bacterial toxins are heat stable, which would allow them to be used in foods that do undergo thermal processing. In 1989, mushrooms were held under conditions, which allowed for the growth and toxin production by Staphylococcus aureus. These mushrooms were subsequently canned, destroying the live bacteria, but allowing the toxin to remain present and cause illness [7]. The mushrooms were grown and canned in another country and then imported into the United States, where the illnesses occurred.

10.2.1.3 Food Processing Systems—Threat Assessment. Food processing may be separated into several distinct phases: production, processing, secondary processing, distribution, sale, and consumption. In addition, transportation is a critical factor at all stages, as food is often produced in one location and then distributed nationally or internationally. Most production sites have minimal security, and are often located in rural settings, where there is a relatively low population density. This leaves the production sites vulnerable to intentional contamination, as most of these locations are easily accessed by a determined individual or group of individuals. Intentional contamination could be carried out at most of these locations with a minimal risk of detection. Voeller V05-c10.tex V1 - 12/04/2013 1:07pm Page 151

SCIENTIFIC OVERVIEW 151

Processing and secondary processing facilities also offer opportunities for intentional contamination to a determined individual or group of individuals. Although most food processing establishments review potential employees prior to hiring, it is quite possible for an individual or individuals with criminal intent to become employed at a food processing facility. There have been several video recordings demonstrating insanitary or inhumane conditions made in secret inside meat processing facilities in the last few years, and it has been suggested that some of these videos may have been made by individuals who sought employment at the facility, specifically to record the video. If an individual can obtain employment with the intent of secretly recording a video, then an individual motivated to deliberately contaminate food could also presumably obtain employment. Virtually, all types of food processing systems incorporate steps during processing to assure the safety of the food product. Although these are designed to address naturally occurring contamination, they are also well suited to address intentional contamination. The more comprehensive food safety systems are based on either the Hazard Analysis Critical Control Point system (HACCP) [8] or the International Organization for Stan- dardization [9]. The disadvantage of these programs is that they are well documented within the processing establishment, and are therefore readily available. A determined individual would be able to evaluate such systems, and devise a way of introducing contamination in a manner that would not allow rapid detection. The food distribution network raises additional concerns. The opportunities for intentional contamination increase, as the food may pass through a series of transportation and storage stages and is potentially available for a variety of individuals to have access to. As with processing, these individuals could conceivably be either employees of the transportation and storage companies or simply determined individuals who gain unauthorized access to the food. A further concern with the distribution network is that most if not all of the microbiological testing of foods is done at the processing level, meaning that foods contaminated during distribution would be unlikely to be tested before they reach the consumer. Foods may be exposed to potential intentional contamination at the point of preparation (food service) or retail sale. At retail, the food is accessible not only by employees of the retail establishment but also by other consumers. It would be very difficult to identify and stop a determined individual from entering a retail establishment and deliberately contaminating food, especially produce that is presented for sale without primary packaging. Previous product tampering cases have involved contamination at the point of sale, and have proved to be difficult to stop. Foods that undergo less processing fall into a higher risk category. Several foods of this type of minimally processed food have been involved in naturally occurring foodborne illnesses in recent years, including lettuce [10] and spinach [11]. Since these types of foods rarely receive additional processing by the consumer before consumption, any pathogen introduced has the potential to cause human illness.

10.2.1.4 Avoidance Strategies. The primary assumption for evaluating a deliberate contamination event is that it would be carried out by an individual or individuals with criminal intent. Because of this, most of the avoidance strategies are based on controlling access of unauthorized individuals to food production and processing areas. Physical security of the food production or processing locations is the ﬁrst step in avoiding a deliberate contamination event. As previously mentioned, food production sites are difﬁcult to Voeller V05-c10.tex V1 - 12/04/2013 1:07pm Page 152

152 MITIGATING CONSEQUENCES OF PATHOGEN INOCULATION INTO PROCESSED FOOD secure, although there are some notable exceptions. Some concentrated animal feeding operations for dairy cattle, poultry, and swine have elaborate physical security measures in place, primarily to prevent the unintentional introduction of animal diseases. These biosecurity measures secure the facility and rigidly control personnel access to the animals. By their nature, locations producing grain crops, leafy green vegetables, or other produce are virtually impossible to secure against unintentional access because of the size of the land mass associated with the growth of these crops. Most food processing and secondary food processing establishments have some degree of physical security. Most are surrounded by perimeter fences and have sufﬁcient exterior lighting to provide for visual examination of the premises. Many employ security guards to regulate access to nonpublic areas of the processing establishment. However, there are practical limitations to all of these measures. There are many people who are not employees of a company, but who have legitimate business within a food processing establishment. These would include, but not be limited to, delivery persons, maintenance and construction personnel, contract pest control and sanitation personnel, customers, and suppliers. Given the intent, any of the individuals could potentially have access to food products. Persons employed by the food production or processing industry have been previously mentioned as potential threats for deliberate contamination. Preemployment screening is vital for many reasons, but even more so to avoid potential contamination. Employees should verify previous employment, and addresses, to assure that the applicant does have a documented history of employment. In addition, preemployment drug screening is widely used to verify that the individual is not currently using illegal drugs. Employees should be adequately trained in the general aspects of food safety, and the impact of a foodborne disease outbreak should be made clear as part of the training. Large foodborne disease outbreaks from natural sources have forced companies to go out of business, resulting in all of the employees becoming unemployed. Emphasizing this point to all employees makes employees more aware, and may result in the reporting of an odd or unusual event witnessed by one of the employees. This empowers employees to become part of the internal monitoring system, and may alert management to an event before the food leaves the processing facility.

10.3 RESPONSE TO AN INCIDENT

The response to an intentional food contamination event would most likely follow a similar pattern as the response to a naturally occurring contamination event. Regret- tably, unless a specific threat has been communicated from the individuals responsible or another method has given some warning, the first indication of a deliberate contamination event would be through public health departments. Standard epidemiological investigations and syndromic surveillance should indicate a common source for the disease outbreak [12]. Once a common source is identified, the response by the food industry should be immediate. The initial response would be to immediately remove all of the food processed in the establishment from commerce, essentially following the practices of a recall. Some of the more specific guidelines for food recalls have been published by the Canadian Food Inspection Agency [13]. The food would then be categorized into specific production lots or code dates to determine which code dates were contaminated. The public health Voeller V05-c10.tex V1 - 12/04/2013 1:07pm Page 153

RESEARCH AND FUNDING DATA 153 investigation should be helpful in identifying whether there were multiple code dates involved, either by recovering intact packages from the homes of the consumer or by investigating the geographical distribution of the disease in relation to the distribution of the food product. This presumes that the food product in question had sufficient lot coding in place to segregate the product. Lot coding becomes more problematic with bulk or commodity items. Extensive microbiological testing would be required to assure that product code dates that are not directly associated with human illnesses were in fact free of the intentional contamination. The deliberately contaminated product must be disposed of in such a way as to assure that it will not pose any additional health or environmental effects. Unlike naturally contaminated product, intentionally contaminated product will likely contain high populations of the pathogen, or high levels of the toxin. Depending on the nature of the contaminant, the food may not be able to be disposed of through the usual methods of rendering, thermal processing, or burying in a land fill. Additional methods may need to be employed to assure that the food poses no further hazard to human health. Com- mercial irradiation, commonly used to sterilize pharmaceutical products and medical devices, may need to be employed to eliminate the contaminant from the food before final disposal. In the case of biological toxins, chemical methods such as those used to destroy animal carcasses infected with bovine spongiform encephalopathy, may have to be employed to ensure the destruction of the toxin. The safety of the workers handling and disposing of the infected products must be considered in the overall plan for disposal. Once the immediate public health crises has subsided, the food industry as a whole should evaluate what happened so that similar incidences may be prevented in the future. Previous historical events have shown that “copycat” events may follow an initial event, and the industry should evaluate the processing systems to prevent a similar incidence from occurring. Naturally occurring contamination events have often identified short- comings in the current processing systems, especially in lot tracing and logistical control of the food product. An intentional contamination event would identify any weaknesses within a given processing system, and an evaluation of the overall event would help to make the industry better prepared for future events.

10.4 RESEARCH AND FUNDING DATA

The type of research needed to address the concerns of deliberate food contamination is similar in many respects to that needed to address naturally occurring contamination. Consequently, the common sources of funding for food safety research are also likely to fund research on intentional contamination, simply because the outcome of the research will be applicable to many situations. The primary government source of research funding for food safety endeavors is the US Department of Agriculture (USDA) Cooperative State Research, Education and Extension Service (CSREES). Major funding programs include the National Research Initiative and the Integrated Research, Education, and Extension Competitive Grants Program. For the updated current request for proposals for these programs, see http://www.csrees.usda.gov/foodsafetybiosecurity.cfm. In addition to the government funding, many of the trade associations offer competitive grants programs in the area of food safety. Although these awards from these grant programs tend to be Voeller V05-c10.tex V1 - 12/04/2013 1:07pm Page 154

154 MITIGATING CONSEQUENCES OF PATHOGEN INOCULATION INTO PROCESSED FOOD smaller than those from the federal programs, they are still a useful source of funding for this type of research.

10.5 CRITICAL NEEDS ANALYSIS

The potential for a deliberate act of contamination to occur is well documented. Histor- ically such events have taken place, although usually perpetrated by an individual intent of ﬁnancial gain or by a disgruntled employee [14]. The potential disruption that could be caused by an organized group, for example for contaminating multiple foods in various locations with different agents, is great. Protecting the food supply from a deliberate contamination event is one of the priorities of the US Department of Homeland, as well as the other agencies that are responsible for the safety of the food supply (USDA Food Safety and Inspection Service (FSIS), Food and Drug Administration (FDA), Environ- mental Protection Agency (EPA), etc.). The food industry is global in its operations, with food being one of the main commodities traded internationally. A deliberate act of contamination in a speciﬁc food could have far reaching impacts not only on human health but also on international trade. A recent incident involving naturally occurring Salmonel- lae contamination of cantaloupe illustrates this point [15]. One of the key components in research is that the outcome of virtually all of the programs will have applications not only for deliberate events but also for contamination that occurs as part of the overall food process.

10.6 RESEARCH DIRECTIONS

Future research needed to address the needs of both regulatory and public health officials in the event of a deliberate event falls into several categories, but can be broadly classified as scientific or technical, logistical, and societal. These are the same research needs for naturally occurring contamination events, so previous, current, and future research programs serve both purposes. The scientific or technical issues relate to sampling and detection methods. Improve- ments need to be made in sampling techniques, as well as the statistical sampling plans needed to determine the extent of contamination. One of the weaknesses of the current detection methods is the lack of an adequate understanding of sampling, with the end result being no better than the quality of the sample collected. Additional resources also need to be applied to detection methodologies. Although significant advances have been made in the area of detection, most notably with the commercialization of polymerase chain reaction (PCR) technology, many of the methods still require an overnight incubation of the sample to increase the number of target cells. In the event of a deliberate contamination event, timing will be essential, and an ideal method would be one that could be performed on-site without a prolonged incubation. Methods that meet these criteria are under development, but have not been entirely successful. The logistical issues that should be addressed include the ability to locate and recover all of the affected food. Recent events, especially with produce and meat [16], have shown how difficult it is to locate and retrieve all food involved. Further research is needed to improve the ability of food processors and federal agencies to track and identify food products to assist in the ability to identify and recall intentionally contaminated food. Voeller V05-c10.tex V1 - 12/04/2013 1:07pm Page 155

REFERENCES 155

Once contaminated food has been recovered, there is an issue of disposal. Depending on the nature of the contaminant, the food may require special processing to render it safe for disposal. In addition, the safety of the workers handling the contaminated food must be considered. As an example, food contaminated with a high population of pathogenic bacteria may require methods of disposal beyond rendering or burying in a land fill. Although it is relatively simple to dispose of small amounts of contaminated food, scale of a potential disposal operation must be considered. The recent actions by USDA in recalling more than 140 million pounds of product [16] illustrate this point. There is a need to study the logistical and technical issues involved in rendering large quantities of food products safe for subsequent disposal. A final research topic would be the effects of an intentional contamination event on society and consumers. An intentional event would raise questions not only about the specific food or foods affected but also about the food supply in general. It would be reasonable and prudent to determine what methods could be used to restore consumer confidence in the general food supply, as well as with a specific commodity. Consumer research could be conducted to determine the most effective means of communicating with consumers after an intentional contamination event, and what messages would be most effective in conveying that the emergency has passed. Presumably, this type of research has been conducted for other natural and intentional catastrophes, and this information could be applied to a food contamination event.

REFERENCES

1. Branigan, W., Allen, M., and Mintz, J. (2004). Tommy Thompson Resigns From HHS . Washing- ton Post.com Dec 3 2004. www.washingtonpost.com/wp-dyn/articles/A31377-2004Dec3.html (accessed 12 September 2006). 2. Torok, T. J., Tauxe, R. V., and Wise, R. P. (1997). A large community outbreak of salmonellosis caused by intentional contamination of restaurant salad bars. JAMA. 278, 389–395. 3. Kolavic, S. A., Kimura, A., Simons, S. L., Slutsker, L., Barth, S., and Haley, C. E. (1997). An outbreak of Shigella dysenteriae type 2 among laboratory workers due to intentional food contamination. JAMA. 278(5), 396–398. 4. Morb. Mortal. Wkly. Rep. (1991). Epidemiologic notes and reports: Cyanide poisonings associated with over-the-counter medication—Washington State, 1991. 40(10), 161, 167–168 5. Wolnik, K. A., Fricke, F. L., Bonnin, E., Gaston, C. M., and Satzger, R. D. (1984). The Tylenol tampering incident—tracing the source. Anal. Chem. 56, 466A–468A, 470A, 474A. 6. Lee, R. V., Harbison, R. D., and Draughon, F. A. (2003). Food as a weapon. Food Prot. Trends. 23(8), 664–674. 7. Levine, W. C., Bennett, R. W., Choi, Y., Henning, K. J., Rager, J. R., Hendricks, K. A., Hopkins, D. P., Gunn, R. A., and Grifﬁn, P. M. (1996). Staphylococcal food poisoning caused by imported canned mushrooms. J. Infect. Dis. 173(5), 1263–1267. 8. NACMCF (1998). Hazard analysis and critical control point principles and application guidelines. J. Food Prot. 61, 1246–1259. 9. ISO (2005). ISO 22000:2005, Food Safety Management Systems - Requirements for Any Orga- nization in the Food Chain. International Organization for Standardization. www.iso.org/ iso/home.htm 10. Morb. Mortal. Wkly. Rep. (2007). Preliminary foodnet data on the incidence of infection with pathogens transmitted commonly through food—10 States, 2006. 56(14), 336–339. Voeller V05-c10.tex V1 - 12/04/2013 1:07pm Page 156

156 MITIGATING CONSEQUENCES OF PATHOGEN INOCULATION INTO PROCESSED FOOD

11. Morb. Mortal. Wkly. Rep. (2006). Ongoing multistate outbreak of Escherichia coli serotype O157:H7 infections associated with consumption of fresh spinach—United States, September 2006. 55(38), 1045–1046. 12. Buehler, J. W., Berkelman, R. L., Hartley, D. M., and Peters, C. J. (2003). Syndromic surveillance and bioterrorism-related epidemics. Emerging Infect. Dis. 9, 1197–1204. 13. CFIA (2007). Food Recalls: Make a Plan and Action It! Manufacturers’ Guide. Updated April 19, 2007 . www.inspection.gc.ca/english/fssa/recarapp/rap/mgguide.shtml (accessed 9 January 2008). 14. Tucker, J. B. (1999). Historical trends related to bioterrorism: an empirical analysis. Emerging Infect. Dis. 5, 498–504. 15. Department of Animal Science (2008). Voluntary Nationwide Recall of Honduran Can- taloupes grown by Agropecuaria Montelibano, San Lorenzo Valle, Honduras, 24 March 2008 . http://www.fda.gov/oc/po/ﬁrmrecalls/centralamerican03 08.html (accessed 7 May 2008) 16. USDA-FSIS (2008). California Firm Recalls Beef Products Derived From Non-Ambulatory Cattle Without the Beneﬁt of Proper Inspection. http://www.fsis.usda.gov/PDF/Recall 005- 2008 Release.pdf (accessed 10 May 2008).

FURTHER READING

(a) AIB (2005). The AIB Guide to Food Security. AIB, Manhattan, KS; (b) National Food Processors Association (2005). Food Security Manual. GMA-NFPA, Washington DC. US Department of Agriculture—Food Safety and Inspection Service. (2008). Food Defense and Emergency Response. http://www.fsis.usda.gov/Food Defense & Emergency Response/ index.asp US Food and Drug Administration. Food Defense and Terrorism. (2008). http://www.cfsan.fda.gov/ ∼dms/defterr.html Voeller v05-bindex.tex V1 - 12/05/2013 1:00pm Page 157

INDEX

aerobiological models, 49 Certiﬁed Crop Advisor (CCA), 46 air contaminants and airborne spread of pathogens, chemical agents, 74–77 123–125 classical swine fever (CSF), 118, 121, 127–129 amperometric biosensors, 11–12 cockroaches angular surface Plasmon resonance biosensing, 9 Homeland Security aspects, 101–102 ants nutrition and development, 100 Homeland Security aspects, 103 pathogens dissemination, 100–101 nutrition and development, 102–103 conductometric biosensors, 13 pathogens dissemination, 103 contaminated foods decontamination and disposal, 69 artiﬁcial insemination, 117 critical needs, 78–79 atomic force microscopy (AFM), 8 fate during disposal biological agents, 73–74 BARC (Bead Array Counter) biosensor, 14 chemical agents, 74–77 base rate fallacy and security, 35 food, 73 biological agents, 73–74 radiological agents, 77–78 biosensors, 5–6 overview for microbial pathogen detection agents, 69–7 electrochemical biosensors, 11–14 decontamination, 72 magnetic biosensors, 14 food, 70 mechanical biosensors, 6–9 research directions, 79–80 optical biosensors, 9–11 Cooperative Agricultural Pest Survey (CAPS), bioterrorism, 43 45–46 and food safety, 89–90 Cooperative State Research, Education and BioWatch, 141, 144 Extension Service (CSREES), 153 counterfeit products and ingredients, 29 CARVER + Shock and Kansas meat industry example, 63–66 Department of Homeland Security (DHS) food safety research and food defense, 66 Customs and Border Inspection (CBI) Program, Centers for Disease Control (CDC), 69, 85, 44, 45 86, 90 direct animal contact and genetic stock Centers for Disease Control and Prevention (CDC), vulnerabilities, vehicles/fomites, and vectors as 1, 3, 4, 140 pathogen sources, 119–121

157 Voeller v05-bindex.tex V1 - 12/05/2013 1:00pm Page 158

158 INDEX

Economic Research Service (ERS), of US Hazard Analysis and Critical Control Point Department of Agriculture (USDA), 3 (HACCP), 28, 62–63, 140, 151 EHEC, 3 hemolytic uremic syndrome (HUS), 3 electrically active polyaniline-coated magnetic high-consequence plant pests early detection and (EAPM) nanoparticles, 14–18 diagnosis, in United States, 43–44 electrochemical biosensors diagnostics amperometric biosensors, 11–12 laboratory throughput, 52–53 conductometric biosensors, 13 US laboratory network system, 53–56 impedimetric biosensors, 13–14 infrastructure, 57 potentiometric biosensors, 12–13 monitoring and surveillance, 44 Environmental Protection Agency (EPA), 70, 74 formal (active) human surveillance, 44–46 enzyme-linked immunosorbent assay (ELISA), 4, 5 nonformal (passive) human surveillance, 46–47 exotic Newcastle disease (END), 125 research, 56–57 risk estimation to orient surveillance, 47–48 feeds and supplements and water vulnerabilities, models, 49–51 121–123 pathway analyses, 48 fluorescence-based biosensors, 10–11 remote and automated surveillance, 51–52 fluorescence resonance energy transfer (FRET), 10 threat analysis, 48 Food and Drug Administration (FDA), 115, 119 highly pathogenic avian influenza (HPAI), 125–126 food-borne threats, 137–140 human illness potential, from animal transmission of food safety, role in food security, 61–62 food-borne pathogens, 133 + CARVER Shock and Kansas meat industry critical needs, 142 example, 63–66 analytical needs, 143 food safety research and food defense, 66 basic science needs, 143–144 food safety education in food security/defense, data needs, 143 66–67 current research, 140 food safety prevention, HACCP, and food control and intervention, 141–142 security/defense, 62–63 detection, 140–141 Food Safety Consortium, 66 research directions, 144 foot-and-mouth disease (FMD), 74, 124, 126–127 scientific overview, 133 foreign animal disease (FAD) and food-borne food-borne threats, 137–140 pathogens farm level control, 111–112 zoonotic threats, 134–137 animal agriculture production in United States, 112 impedimetric biosensors, 13–14 US beef industry, 113–114 insects, as vectors of foodborne pathogens, 97 US pork industry, 117–119 US poultry industry, 114–116 ants current critical important pathogens for Homeland Security aspects, 103 food-producing industries, 125 nutrition and development, 102–103 classical swine fever, 127–129 pathogens dissemination, 103 foot-and mouth disease, 126–127 cockroaches highly pathogenic avian influenza (HPAI), Homeland Security aspects, 101–102 125–126 nutrition and development, 100 disease introduction at farm level preventing and pathogens dissemination, 100–101 controlling, 119 muscoid flies and fruit flies, 97–98 direct animal contact and genetic stock nutrition and development, 98 vulnerabilities, vehicles/fomites, and Homeland Security aspects, 99 vectors as pathogen sources, 119–121 pathogens dissemination and antibiotic resistant air contaminants and airborne spread of strains, 98–99 pathogens, 123–125 pantry pests feeds and supplements and water Homeland Security aspects, 104 vulnerabilities, 121–123 nutrition and development, 103 vaccine and pharmaceutical vulnerabilities, 123 pathogens dissemination, 104 formal (active) human surveillance, 44–46 integrated extraction/detection magnetic nanoparticle-based biosensor system, 14–19 giant magnetoresistive (GMR) sensors, 14 integrated pest management Pest Information Platform for Extension and Education Hall-effect, 14 (ipmPIPE), 46, 47 Voeller v05-bindex.tex V1 - 12/05/2013 1:00pm Page 159

INDEX 159

Integrated Research, Education, and Extension Nernst equation, 12 Competitive Grants Program, 153 nonformal (passive) human surveillance, 46–47 International Organization for Standardization, 151 Nuclear Regulatory Commission, 77 ion selective filed effect transistors (ISFETs), 12 nucleic acid sequence based amplification (NASBA), 11 light addressable potentiometric sensor (LAPS), 12–13 Occupational Safety and Health Administration low pathogenic avian influenza (LPAI), 125, 126 (OSHA), 69 OIE (World Organization for Animal Health), 127 magnetic biosensors, 14 operational risk management (ORM), 28, 31, mechanical biosensors 35, 38 microcantilever-based biosensors, 8–9 optical biosensors quartz crystal microbalance (QCM) biosensors, fluorescence-based biosensors, 10–11 6–7 surface Plasmon resonance biosensors (SPR), surface acoustic wave (SAW) biosensors, 7–8 9–10 microbial culturing, 4 over-the-counter (OTC) drugs, 30, 31 microbial risk assessment (MRA), 142 microbiological detectors, for food safety packaging, 27–28, 29–30. See also system applications flexibility and response biosecurity and food safety threats, 1–4 add-on indicators, 32–33 biosensors for microbial pathogen detection multiple authentication, 35 electrochemical biosensors, 11–14 optical systems, 34 magnetic biosensors, 14 physical token systems and RFID, 34 mechanical biosensors, 6–9 proactive devices, 33 optical biosensors, 9–11 product authentication, 34 detection, 4–6 and safety assurance, 30–31 integrated extraction/detection magnetic security and base rate fallacy, 35 nanoparticle-based biosensor system, 14–19 tamper evidence requirements, 31 microcantilever-based biosensors, 8–9 tamper indication devices, 31–32 Model for Organic Chemicals in Landfills pantry pests (MOCLA), 74, 75, 76 Homeland Security aspects, 104 multi-locus variable number tandem repeat analysis nutrition and development, 103 (MLVA), 91, 92 pathogens dissemination, 104 muscoid flies and fruit flies, 97–98 pathogen inoculation into processed food and mitigating consequences, 149 nanoparticles, 14–19 critical needs analysis, 154 National Agricultural Pest Information System, 46 research and funding data, 153–154 National Animal Health Monitoring System research directions, 154–155 (NAHMS), 118, 119 response to incident, 152–153 National Center for Animal Health Emergency scientific overview Management (NCAHEM), 127 processed foods, 149–152 National Center for Food Protection and Defense, Peanut Corporation of America (PCA), 37–38 66 physical token systems and RFID, 34 National Center of Excellence for Foreign Animal polymerase chain reaction (PCR), 4, 5 and Zoonotic Disease Defense, 142 population dynamics models, 50–51 National Institute of Allergy and Infectious Diseases potentiometric biosensors, 12–13 (NIAID), 1, 3 preemployment screening, 152 National Oceanic and Atmospheric Administration processed foods and pathogens (NOAA) avoidance strategies, 151–152 HYSPLIT (hybrid single-particle Lagrangian historical precedence, 149–150 integrated trajectory), 49, 50 potential biological agents, 150 National Plant Diagnostic Network (NPDN), 47, 51, system, and threat assessment, 150–151 53–56 processing, 27–29. See also system flexibility and National Poultry Improvement Plan (NPIP), response 125–126 counterfeit products and ingredients, 29 National Research Initiative, 153 product authentication, 34 National Veterinary Services Laboratories (NVSL), pulsed-field gel electrophoresis (PFGE), 85, 86, 88, 126 89, 91 Voeller v05-bindex.tex V1 - 12/05/2013 1:00pm Page 160

160 INDEX

PulseNet, 83–84 syndromic analysis, 51 PulseNet International, 87 system flexibility and response, 36 critical need, 89–90 cascading failure in food processing and structure and function, 87–89 packaging system, 36–37 PulseNet USA, 85–87 safety system failure case study, 37–38 research directions, 91–92 scientific overview to detect and track food thrombotic thrombocytopenic purpura (TTP), 3 contamination, 84–85 PulseNet Canada, 88 United States Department of Agriculture (USDA), PulseNet International, 84, 87, 88 46, 111, 116, 125, 134, 153 critical need, 89–90 Animal and Plant Health Inspection Service structure and function, 87–89 (APHIS), 44, 45, 48 PulseNet International ListServ, 89 Universal Product Code (UPC), 34 PulseNet USA, 85–87 US beef industry, 113–114 US Food and Drug Administration, 61 quartz crystal microbalance (QCM) biosensors, US laboratory network system, 53–56 6–7 US National Research Council, 43 US pork industry, 117 radiological agents, 71, 77–78 farrow-to-wean, 117–118 RAPTOR™,10 grower and finisher site, 118–119 remote and automated surveillance, 51–52 nursery, 118 reproductive number (R), 50 US poultry industry, 114–115 Rift Valley fever (RVF) virus, 134, 136, 137 typical poultry company, 115–116 safety system failure case study, 37–38 vaccine and pharmaceutical vulnerabilities, 123 Sauerbrey equation, 6 Sewage Treatment Plant Fugacity Model (STPWIN), wastewater treatment plants (WWTPs), 72–78 76 weather-based GIS and disease/pest warning models, single nucleotide polymorphisms (SNPs), 91, 92 49–50 surface acoustic wave (SAW) biosensors, 7–8 West Nile virus (WNV), 134, 136 surface Plasmon resonance biosensors (SPR), 9–10 surge management, 53 zoonotic threats, 134–137