Thermal Inactivation of Escherichia coli O157:H7 in Blade-Tenderized Beef Steaks Cooked on a Commercial Open-Flame Gas Grill

1404 Journal of Food Protection, Vol. 72, No. 7, 2009, Pages 1404 1411 Thermal Inactivation of Escherichia coli O157:H7 in Blade-Tenderized Beef Steaks Cooked on a Commercial Open-Flame Gas Grill JOHN B. LUCHANSKY, 1 * ANNA C. S. PORTO-FETT, 1 BRADLEY SHOYER, 1 RANDALL K. PHEBUS, 2 HARSHAVARDHAN THIPPAREDDI, 3 AND JEFFREY E. CALL 1 1 U.S. Department of Agriculture, Agricultural Research Service, Eastern Regional Research Center, Microbial Food Safety Research Unit, 600 East Mermaid Lane, Wyndmoor, Pennsylvania 19038; 2 Department of Animal Sciences and Industry, Kansas State University, Manhattan, Kansas 66506; and 3 Department of Food Science and Technology, University of Nebraska, Lincoln, Nebraska 68583, USA MS 08-618: Received 19 December 2008/Accepted 12 February 2009 ABSTRACT Beef subprimals were inoculated on the lean side with ca. 4.0 log CFU/g of a cocktail of rifampin-resistant (Rif r ) Escherichia coli O157:H7 strains and then passed once through a mechanical blade tenderizer with the lean side facing upward. Inoculated subprimals that were not tenderized served as controls. Two core samples were removed from each of three tenderized subprimals and cut into six consecutive segments starting from the inoculated side. A total of six cores were also obtained from control subprimals, but only segment 1 (topmost) was sampled. Levels of E. coli O157:H7 recovered from segment 1 were 3.81 log CFU/g for the control subprimals and 3.36 log CFU/g for tenderized subprimals. The percentage of cells recovered in segment 2 was ca. 25-fold lower than levels recovered from segment 1, but E. coli O157:H7 was recovered from all six segments of the cores obtained from tenderized subprimals. In phase II, lean-side inoculated (ca. 4.0 log CFU/g), single-pass tenderized subprimals were cut into steaks of various thicknesses (1.91 cm [0.75 in.], 2.54 cm [1.0 in.], and 3.18 cm [1.25 in.]) that were subsequently cooked on a commercial open-flame gas grill to internal temperatures of 48.8 C (120 F), 54.4 C (130 F), and 60 C (140 F). In general, regardless of temperature or thickness, we observed about a 2.6- to 4.2-log CFU/g reduction in pathogen levels following cooking. These data validate that cooking on a commercial gas grill is effective at eliminating relatively low levels of the pathogen that may be distributed throughout a blade-tenderized steak. There is little debate that Escherichia coli O157:H7 is a significant threat to public health and/or that undercooked meat is an important vehicle for human illness (7). In 2008 alone, the U.S. Department of Agriculture/Food Safety and Inspection Service (USDA/FSIS) called for the voluntary recall of ca. 68 million kg of raw beef due to possible contamination with E. coli O157:H7 (http://www.fsis. usda.gov/fsis Recalls) resulting in at least two outbreaks that have been responsible for at least 105 illnesses (http://www.cdc.gov; www.ift.org/cgi-bin/news). It should be noted, however, that produce, dairy products, and water have also been responsible for human illness and rather large recalls due to E. coli O157:H7 in recent years (3, 4, 10, 25). Thus, appreciable resources are being directed to better manage this pathogen in various commodities both pre- and postharvest. Although undercooked ground beef is arguably the pri- * Author for correspondence. Tel: 215-233-6676; Fax: 215-233-6581; E-mail: John.Luchansky@ars.usda.gov. Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the U.S. Department of Agriculture. Portions of this research were presented at the 21st Meeting of the International Committee on Food Microbiology and Hygiene, 1 to 4 September 2008, Aberdeen, Scotland. mary vehicle for transmission of E. coli O157:H7, in recent years there have been at least four outbreaks of E. coli O157:H7 epidemiologically associated with the consumption of nonintact, mechanically and/or chemically tenderized steaks, sometimes also referred to as injected-tenderized steaks (1, 2, 15, 31). Products categorized as such include almost any whole or intact and comminuted meats subsequently exposed to mechanical or chemical enhancement with solid or hollow needle injectors or blades (28). Different approaches and/or technologies have been used to tenderize meats, such as chemical injection, enzymatic digestion, hydrodynamic shock, and/or blade and needle tenderization (17, 28). In so doing, the normally sterile internal tissues of a previously intact piece of meat may become contaminated with microbes from the nonsterile external surface of the meats, some of which may include foodborne pathogens. Even if tenderization can transfer pathogens into the deeper tissues, the likelihood of this occurring is rare owing to the low incidence of E. coli O157:H7 on the surface of subprimals. For example, a 2004 survey (14) reported that E. coli O157:H7 was not detected on the surface of any of 1,199 intact subprimals tested (incidence of 0.083%). These data are similar to those reported in an earlier study by Warren (32) wherein ca. 0.2% (2 of 1,014 samples) of subprimal beef cuts surveyed were surface con-

J. Food Prot., Vol. 72, No. 7 FATE OF E. COLI O157:H7 IN COOKED NONINTACT BEEF 1405 FIGURE 2. Photograph showing the approximate location from which each of the two core samples were taken from tenderized subprimals (A) and the stainless steel device used to extract core samples from each subprimal (B). FIGURE 1. Photograph of a beef subprimal entering the blade tenderizer with the lean, inoculated side facing upward. taminated with E. coli O157:H7 at levels of 0.375 CFU/ cm 2. That being said, results from the handful of scientifically controlled studies published to date indicate that relatively few cells are translocated from the surface into the deeper tissues of the meat following tenderization (2, 28). For example, Gill and colleagues (8, 9) reported that aerobes, coliforms, and listeriae/staphylococci were recovered from the internal portions of meat tenderized both at a packing plant and at retail. Lambert and colleagues (16) reported that, depending on the method for tenderization, between 1 and 8% of salmonellae were internalized into pork. As other examples, Luchansky et al. (19) and Phebus et al. (24) reported that most E. coli O157:H7 cells remained in the topmost 1 cm and that 3 to 4% were translocated to the geometric center of blade-tenderized subprimals, respectively. Based on these and related data, the National Advisory Committee on Microbiological Criteria for Foods (22) concluded that nonintact meat does not pose a greater risk to public health than intact beef steaks, provided that the meat is cooked to an internal temperature of 60 C. This same expert body also called for additional research related to blade-tenderized meats that included the need for further experiments to quantify survival of E. coli O157: H7 in core beef samples following cooking to specified temperatures (22). In 1994, the USDA/FSIS issued a policy stating that raw ground beef is to be considered adulterated when contaminated with E. coli O157:H7, and in 1999, the Agency extended its adulteration policy to include all raw, nonintact beef contaminated with this pathogen (30). Despite the association of foodborne illnesses with the consumption of raw or undercooked meat contaminated with E. coli O157: H7, Schmidt et al. (27) reported that ca. 30% of consumers (63 of 210 surveyed) ordered their steaks cooked medium rare (internal temperature of ca. 60.0 to 62.8 C [140 to 145 F]) and that 10% of consumers (21 of 210 surveyed) ordered their steaks rare (internal temperature of ca. 54.4 to 57.2 C [130 to 135 F]) or very rare (internal temperature of ca. 48.8 to 51.6 C [120 to 125 F]). Cox et al. (5) also reported that ca. 53% of consumers (1,828 of 3,424 surveyed) ordered their steaks cooked medium to medium rare, whereas ca. 4.5% of consumers (150 of 3,424) ordered steaks cooked rare. Although up to 18% of beef products sold at retail are mechanically tenderized and/or enhanced (1), there is relatively little information available concerning safe cooking guidelines that would eliminate internalized E. coli O157:H7 in such nonintact products. Thus, the primary objective of this study was to determine if commonly used cooking temperatures are lethal for E. coli O157:H7 translocated into steaks cut from mechanically tenderized subprimals. MATERIALS AND METHODS Bacterial strains. The five strains making up the rifampinresistant (Rif r ) E. coli O157:H7 cocktail (USDA/FSIS 011-82, ATCC 43888, ATCC 43889, ATCC 43890, and USDA/FSIS 45756) used in this study to inoculate the surface of beef subprimals were confirmed, cultured, and maintained as described previously (19). In brief, a single colony of each isolate was separately grown in tryptic soy broth (Difco, Becton Dickinson, Sparks, MD) overnight at 37 C without shaking. The 10-ml volumes (one for each freshly grown culture) were combined, and the resulting suspension was harvested by centrifugation and resuspended in 50 ml of sterile 0.1% peptone water (Difco, Becton Dickinson) before being diluted as appropriate in peptone water to attain the desired target inoculum level. Inoculation and tenderization of subprimals. Top sirloin butt beef subprimals (USDA Institutional Meat Purchase Specifications no. 184; ca. 6.8 to 9 kg each) were purchased from a local wholesale distributor and stored at 4 C for up to 12 days. In our previous study (19), we reported that greater translocation was observed following lean-side inoculation and single-pass tenderization than following double-pass tenderization and/or following fat-side inoculation and single- or double-pass tenderization. As such, in the present study, the lean-side surface of each top butt beef subprimal was inoculated with 10 ml (ca. 4.0 log CFU/g) of the above-mentioned Rif r E. coli O157:H7 cocktail following removal of the subprimals from the original vacuum-sealed packages. For each of two trials, one subprimal was not tenderized (control), whereas an otherwise similar set of three subprimals were single-pass tenderized and subsequently evaluated for the extent of E. coli O157:H7 translocation by using the coring and plating methods that we recently published (Figs. 1 to 3) (19). As described in greater detail below, the three tenderized subprimals were cut into steaks, some of which were subsequently cooked to

1406 LUCHANSKY ET AL. J. Food Prot., Vol. 72, No. 7 FIGURE 4. Tenderized subprimals being cut into steaks with a specially constructed plastic mitre box. FIGURE 3. Photograph of core samples showing the approximate location of each of the six segments that were obtained. Segments 1 to 4 were ca. 1 cm thick, whereas segments 5 and 6 were ca. 2 cm thick. three target temperatures on a gas grill. The tenderization component herein consisted of 3 subprimals 2 core samples per each subprimal 6 segments per each core sample 2 trials for a total of 72 samples from tenderized subprimals. Since only segment 1 of the nontenderized subprimal was sampled, 12 samples were tested from control (nontenderized) subprimals (1 subprimal 6 core samples per each subprimal 1 segment per each core sample 2 trials). Cooking steaks. Tenderized subprimals were cut into steaks ca. 1.91, 2.54, and 3.18 cm thick (0.75, 1.0, and 1.25 in., respectively) by using a specially constructed plastic mitre box (Fig. 4). A total of three steaks of each thickness were cooked on an openflame commercial gas grill (model XXE-4, Baker s Pride, New Rochelle, NY [Fig. 5]) using natural gas, and with all 4 burners fully open to achieve target internal temperatures of 48.9 C (120 F), 54.4 C (130 F), and 60 C (140 F) (Fig. 5). The steaks were flipped over at the approximate midpoint between the initial and target endpoint temperatures in an effort to cook both sides of each steak equally and achieve the desired internal temperature of the meat. The internal temperature of each steak was monitored using two calibrated stainless steel, type J thermocouples (Omega Engineering Inc., Stamford, CT), each of which was inserted into opposing sides to the approximate geometric center of each steak. The thermocouples from each steak were individually connected to a six-channel digital panel temperature indicator (model 500T, Doric Instruments, VAS Engineering Inc., San Diego, CA), and the temperature was monitored visually until the target internal temperature of the meat was achieved. Steaks were then removed from the grill with a sterile spatula, transferred onto sterile polystyrene foam packing trays (Koch Supplies, Kansas City, MO), and immediately portioned into strips and quarters as described below. The cooking component of this study consisted of 3 cooking temperatures 3 thicknesses of steaks 3 steaks per each thickness 2 trials for a total of 54 cooked steaks (9 steaks for each cooking temperature per each of two trials) cut from tenderized subprimals. The controls (uncooked steaks cut from tenderized subprimals) consisted of 3 thicknesses of steaks 3 steaks per each thickness 2 trials for a total of 18 uncooked steaks. Microbiological analyses. Cells of E. coli O157:H7 were recovered from tenderized beef samples as described previously (19), whereas cells surviving after cooking were recovered as follows. Briefly, both cooked and noncooked steaks were portioned (Fig. 6) into strips (S1, S2, and S3) and quarters (Q1, Q2, Q3, and Q4), which were separately weighed and subsequently blended and macerated for ca. 30 s (Magic Bullet, http://www. healthandfitnessproducts.com). The resulting fluid was plated, with and without prior dilution in sterile 0.1% peptone water, onto sorbitol-macconkey agar plus rifampin (100 g/ml; Sigma Chemical Company, St. Louis, MO). Plates were incubated at 35 C for 24 h, and sorbitol-negative colonies were enumerated as log CFU of E. coli O157:H7 per gram. When pathogen levels decreased to below the detection limit ( 0.85 log CFU/g for strips and 0.90 log CFU/g for quarters) by direct plating, the strips or quarters that tested negative for the pathogen by direct plating were separately composited and enriched as previously described (19). Statistical analyses. For phase I of the study, as described previously (19), transfer of E. coli O157:H7 cells into deeper tissues of subprimals by way of mechanical tenderization was expressed (as a percentage) as the number of cells (CFU per gram) recovered separately from each of the six segments obtained from tenderized subprimal cores, divided separately by the number of cells (CFU per gram) recovered from segment 1 of the cores obtained from the nontenderized, positive control subprimals. The standard deviations for the levels of the pathogen recovered from each of the six segments and the cumulative totals recovered from core samples were calculated using the statistical function option that is provided with Microsoft Excel 2003 software (Redmond, WA). For phase II of this study, the SAS system (version 8.0, SAS Institute, Cary, NC) was used to determine statistically significant differences among steak thicknesses, cooking temperatures, and sample types (i.e., strips versus quarters). Means and standard deviations in the cooking experiments were calculated from individual sets of data for each of the two separate trials at each of the three temperatures tested using triplicate samples at each time interval. Analysis of variance was used to determine the effects and interactions of the factors on the log reduction values. Differences in lethality observed for each temperature, thickness, sample type, and/or combinations thereof were considered significant for P values of 0.05. RESULTS AND DISCUSSION Translocation of E. coli O157:H7 into subprimals via blade tenderization. The epidemiological association of nonintact beef with human illness in recent years has

J. Food Prot., Vol. 72, No. 7 FATE OF E. COLI O157:H7 IN COOKED NONINTACT BEEF 1407 TABLE 1. Levels of E. coli O157:H7 recovered from segmented core samples obtained from single-pass tenderized subprimals Segmented core sample log CFU/g (SD) % transfer a Range of transfer within segment (log CFU/g) FIGURE 5. Tenderized steak being cooked on an open-flame commercial grill. Note the weighted flat iron (ca. 1.4 kg) used to prevent curling/cupping of steaks on the grill during cooking and the thermocouples inserted into the approximate geometric center of the steak. Positive control 3.89 (0.28) NA 2.89 4.30 Experimental tenderized cores (segment nos.) 1 3.40 (2.03) 33 1.60 3.89 2 2.37 (1.05) 3.04 0.90 3.45 3 2.38 (0.96) 3.09 0.90 3.80 4 1.53 (0.81) 0.44 0.90 2.70 5 0.91 (0.64) 0.10 0.90 1.81 6 1.80 (0.86) 0.82 0.90 3.05 a Percent transfer is calculated as (CFU/g of tenderized subprimal core segment/cfu/g of segment 1 of the nontenderized control subprimal core) 100. NA, not applicable. renewed concerns among risk assessors and policy makers that it may present a risk to public health akin to the risk presented by other nonintact products, notably ground beef, if it is undercooked and then consumed. There have only been a handful of scientifically valid studies that demonstrated, and in some instances quantified, translocation of E. coli O157:H7 from the surface into the deeper tissues of whole muscle meat as a consequence of blade tenderization. Regarding the latter, Phebus et al. (24) were the first to report that 3 to 4% of cells were translocated into the geometric center of beef subprimals following tenderization. In collaboration with these investigators, we recently confirmed that cells of E. coli O157:H7 are indeed translocated into the geometric center of lean-side inoculated and tenderized steaks, and we also quantified for the first time that the majority of the cells remained in the topmost 1 cm of a tenderized beef subprimal (19). The testing of six segments from each of 10 core samples from each tenderized subprimal allowed us to quantify that, depending on the initial inoculum, between 32 and 41% of cells remained in the top segment. Similar results were obtained in the present study, wherein the majority (3.4 log CFU/g; 33%) of the 3.81 log CFU/g of lean-side inoculated E. coli O157: H7 was transferred into segment 1 (Table 1). Lesser levels of the pathogen were recovered from segment 2 (2.37 log CFU/g; 3.04%), segment 3 (2.38 log CFU/g; 3.09%), segment 4 (1.53 log CFU/g; 0.44%), segment 5 (0.91 log CFU/g; 0.10%), and segment 6 (1.80 log CFU/g; 0.82%). The summation of cells recovered from all six segments from all core samples was ca. 40%. These results are consistent with our previous work wherein we reported that regardless of inoculation level (ca. 0.5 to 3.5 log CFU/g) total levels recovered from all six segments of a core sample were 37 to 55% (20). The results of the present study (1.53 log CFU/g [0.44%] recovered in segment 4, the approximate geometric center) are somewhat lower in absolute value than the results of our previous work (3 to 4% in geometric center (19, 24, 29)) but nonetheless compare favorably in that relatively few cells are carried into the deeper interior portions of the meat. Thermal inactivation of E. coli O157:H7 in nonintact steaks cooked on a gas grill. Of the various interventions used by both the industry and consumers to control pathogens in raw meat, cooking is arguably the most effective and most readily available and economically feasible method. Cooking raw meat at the proper temperature for the proper time is generally quite effective for killing vegetative cells of the microbial pathogens responsible for causing most foodborne illness episodes (12, 23, 26). As discussed in greater detail below, there have been numerous published studies on thermal inactivation of E. coli O157: H7 in both ground and whole muscle meats, whereas far less has been published on thermal inactivation of this pathogen and others in nonintact beef. In the second phase of the present study, we sought to quantify the lethality achieved by cooking blade-tenderized steaks containing E. coli O157:H7 on a commercial gas grill (Fig. 5). As for the tenderization component of the present study, the manner in which the meat was portioned for sampling was critical for quantifying the total kill throughout each steak. With the knowledge that most cells were likely to reside in the topmost 1 cm (19), that being the lean side, outer rim of the steak, as shown in Figure 6, we carefully dissected three strips (S1, S2, and S3) from segment 1, each of which was ca. 1 cm in width, ca. 21 cm long, and either ca. 1.91, 2.54, or 3.18 cm thick (precooking thickness). The remainder of each steak was portioned into four quarters (Q1, Q2, Q3, and Q4). In general, appreciably more cells were recovered from uncooked strips than from uncooked quarters, as expected, since the latter comprised the topmost 1 cm of the tenderized subprimals wherein most of the cells were translocated (Table 2). For a given cooking temperature, there were no statistical differences (P 0.05) in the extent of thermal inactivation of E. coli O157:H7 that were transferred via blade tenderization into strips (topmost 1 cm; S1, S2, and

1408 LUCHANSKY ET AL. J. Food Prot., Vol. 72, No. 7 FIGURE 6. Uncooked (A and B) and cooked (C and D) tenderized steak portioned into strips (S1, S2, and S3) and quarters (Q1, Q2, Q3, and Q4). S3) from steaks cut to a thickness of either 1.91, 2.54, and 3.18 cm (Table 2). However, inactivation of E. coli O157: H7 was greater (P 0.05) on strips cut from 1.91-cm-thick steaks cooked to a target internal temperature of 48.9 C than on otherwise similar strips and steaks cooked to a target internal temperature of 60 C, but not 54.4 C. No statistical differences (P 0.05) in lethality were observed among strips cut from steaks that were 2.54 or 3.18 cm thick and cooked to target internal temperatures of 48.9, 54.4, or 60 C. Knowing that some cells would also reside in the geometric center of the steaks, we carefully quartered (Q1, Q2, Q3, and Q4) the remainder of each cooked steak so that ultimately the entirety of each steak was sampled, albeit by taking the summation of strips 1, 2, and 3 plus the summation of quarters 1, 2, 3, and 4. In general, regardless of the thickness of the steaks or the temperatures at which they were cooked, no statistical differences (P 0.05) were observed among the three strips or among the four quarters relative to inactivation of E. coli O157:H7 in mechanically tenderized steaks (Table 2). Furthermore, for a given cooking temperature and thickness of the steak, with the exception of steaks that were cut to a thickness of 2.54 or 3.18 cm and cooked to target internal temperatures of 48.9 or 54.4 C, there were no statistical differences (P 0.05) in the extent of thermal inactivation of E. coli O157:H7 cells that were transferred via blade tenderization between the summation of strips (topmost 1 cm; S1, S2, and S3) and the summation of quarters (Q1, Q2, Q3, and Q4) (Table 2). Likewise, regardless of the cooking temperature or the thickness of the steak, in general, there were TABLE 2. Levels of E. coli O157:H7 recovered from nonintact steaks before and after cooking a log CFU/g SD Temp ( C) Thickness (cm) Uncooked Strips (S1 S2 S3) Cooked Quarters (Q1 Q2 Q3 Q4) Uncooked Cooked Total steak (all strips all quarters) b Uncooked Cooked 48.9 1.91 3.70 0.17 1.04 0.11 a B 2.78 0.31 0.89 0.39 a A 3.76 0.15 1.36 0.15 a A 2.54 3.71 0.19 0.55 0.17 a A 2.51 0.23 0.85 0.50 a A 3.75 0.17 1.02 0.38 a A 3.18 4.07 0.29 0.58 0.29 a A 2.69 0.22 0.74 0.36 a A 4.09 0.28 0.91 0.38 a A 54.4 1.91 3.93 0.29 0.92 0.49 a AB 3.12 0.32 0.85 0.38 a A 3.85 0.29 1.24 0.48 a A 2.54 3.81 0.10 0.73 0.40 a A 2.69 0.49 0.68 0.42 a A 3.94 0.11 0.97 0.48 a A 3.18 3.89 0.12 0.37 0.00 a A 2.81 0.32 0.67 0.47 a A 4.20 0.11 0.66 0.45 a A 60 1.91 3.83 0.51 0.37 0.00 a A 3.10 0.96 0.92 0.43 a A 3.95 0.63 0.93 0.44 a A 2.54 4.15 0.80 0.64 0.42 a A 3.24 0.94 0.79 0.46 a A 4.23 0.83 0.93 0.46 a A 3.18 3.94 0.57 0.79 0.39 a A 3.13 0.67 0.71 0.44 a A 4.00 0.58 1.03 0.43 a A a For a given temperature, thickness means with different lowercase letters in common within a column are significantly (P 0.05) different by the Bonferroni least significant difference (LSD) test; for a given thickness, temperature means with different uppercase letters in common within a column are significantly (P 0.05) different by the Bonferroni LSD test; for a given thickness and temperature, means with different lowercase Greek letters in common within rows are significantly (P 0.05) different by the Bonferroni LSD test. b Values for Total steak are the summation of total CFU from all strips plus all quarters and represent the results from two trials and 42 pieces of meat.

J. Food Prot., Vol. 72, No. 7 FATE OF E. COLI O157:H7 IN COOKED NONINTACT BEEF 1409 TABLE 3. Reductions of E. coli O157:H7 from nonintact steaks that were segmented after cooking a Temp ( C) Thickness (cm) log CFU/g SD Strip 1 Strip 2 Strip 3 Quarter 1 Quarter 2 Quarter 3 Quarter 4 48.9 1.91 2.88 0.37 A 3.24 1.15 A 3.56 1.94 A 1.57 1.03 A 3.05 0.75 A 2.50 2.02 A 3.94 1.31 A 2.54 4.11 0.94 A 4.00 1.18 A 4.14 1.65 A 2.95 1.47 A 2.21 0.94 A 2.38 0.76 A 1.70 0.89 A 3.18 4.72 1.32 A 4.76 0.37 A 4.51 0.83 A 3.17 1.37 AB 3.04 1.10 AB 2.15 1.05 AB 0.96 0.98 B 54.4 1.91 3.58 1.68 A 3.68 1.32 A 4.32 1.15 A 3.10 1.34 A 2.69 0.92 A 2.50 1.15 A 1.66 1.03 A 2.54 3.56 0.95 A 3.83 1.29 A 3.89 0.93 A 2.37 1.58 A 1.98 2.13 A 2.51 1.46 A 2.77 1.86 A 3.18 4.57 0.16 A 4.77 0.38 A 4.71 0.25 A 2.44 1.52 A 3.53 1.36 A 2.58 1.51 A 2.55 1.25 A 60 1.91 4.61 1.55 A 4.47 0.76 AB 4.77 0.98 A 2.66 1.38 AB 2.27 1.27 AB 1.38 1.88 B 2.73 1.82 AB 2.54 4.30 1.61 A 4.27 1.39 A 4.45 1.69 A 2.77 1.48 A 2.91 1.13 A 2.73 1.82 A 1.88 1.67 A 3.18 3.60 1.19 A 3.96 0.96 A 4.22 1.22 A 3.09 1.32 A 3.61 1.48 A 1.84 2.01 A 2.53 1.02 A a For a given temperature and thickness, means with different letters within a row are significantly (P 0.05) different by the Bonferroni least significant difference test. TABLE 4. Pathogen recovery by direct plating and enrichment of cooked steak portions testing negative for E. coli O157:H7 Temp ( C) Thickness (cm) No. of positive samples/total no. Strips (1 2 3) a Direct plating Quarters (1 2 3 4) b Direct plating Enrichment Enrichment 48.9 1.91 4/18 5/14 3/24 7/21 2.54 0/18 5/18 4/24 3/20 3.18 3/18 2/15 6/24 1/18 54.4 1.91 1/18 3/17 1/24 4/23 2.54 0/18 2/18 1/24 3/23 3.18 0/18 0/18 1/24 1/23 60 1.91 1/18 1/17 0/18 0/18 2.54 0/18 2/18 0/24 3/24 3.18 2/18 2/16 2/24 1/22 a Results for a composite of strips 1, 2, and/or 3 (summation of 3 steaks 3 strips 2 trials; 18 strips total per each temperature) obtained from cooked steaks. b Results for a composite of quarters 1, 2, 3, and/or 4 (summation of 3 steaks 4 quarters 2 trials; 24 quarters total per each temperature) obtained from cooked steaks. no statistical differences (P 0.05) in inactivation of E. coli O157:H7 between individual strips and individual quarters (Table 3). There was, however, significantly (P 0.05) less lethality within Q4 of steaks cut to a thickness of 3.18 cm and cooked to 48.9 C than in the corresponding S1, S2, and S3 from these same steaks, but not compared with Q1, Q2, or Q3 (Table 3). Similarly, there was significantly (P 0.05) more lethality in S1 and S3 for 1.91- cm-thick steaks cooked to 60 C than in the corresponding Q3 for these same steaks. That being said, cooking to 48.9, 54.4, and 60 C reduced E. coli O157:H7 levels in both strips and quarters of tenderized steaks by ca 2.6 to 4.2 log CFU/g (Table 2). However, with two exceptions, those being for strips cut from 3.18-cm-thick steaks cooked to 54.4 C and quarters cut from 1.91-cm-thick steaks cooked to 60 C, it was possible to recover the pathogen via enrichment after cooking in all segments and all quarters, regardless of temperature or thickness (Table 4). Our data compare favorably with those of Sporing (29), whereby E. coli O157:H7 reductions ranging from 2.4 to 4.8 log CFU/g were achieved after cooking mechanically tenderized steaks (1.27 to 3.81 cm thick) to 54.4 or 60 C on an open-flame gas grill. Although not readily apparent in the present study, it was anticipated that lethality should increase appreciably with increasing cooking temperatures from 48.9 to 71.1 C (120 to 160 F). To observe and/or quantify this, it will be necessary to use higher initial levels of E. coli O157:H7 so that the total reductions can be measured. Such studies are now in progress. As reported by other investigators (12, 23), higher fat levels in foods afford pathogens greater protection from heat, probably due to the reduced water activity and/or the reduced heat penetration (lower heat conductivities) through the heating menstruum. Although the fat content of the subprimals and nonintact steaks used in this study was not directly measured, in general, the fat content of cooked steaks can range from 3.3 to 7.7% (18, 21). For beef steaks, the fat is distributed primarily around the outer rim of the fat side and via marbling rather than being more uniformly distributed as for ground beef. In general, and depending on the fat content, cooking parameters, and grill type, other investigators reported reductions of E. coli O157:H7 in red meat, particularly ground beef patties, ranging from ca. 1.5 to 5.5 log CFU/g when cooked to internal temperatures of 60 or 68.3 C (140 or 155 F) (6, 11, 13). As a general comparison, these data compare favorably with the ca. 2.6- to 4.2-log CFU/g reductions we observed following cooking of nonintact, needle-tenderized steaks containing E. coli O157:H7 to an internal target temperature of 48.9, 54.4, or 60 C on an open-flame gas grill. In conclusion, blade tenderization distributed E. coli O157:H7 throughout a beef subprimal, but appreciably more cells were transferred into the topmost 1 cm than into the deeper tissues and segments of the resulting nonintact beef. Regardless of steak thickness or whether more cells were distributed on the surface (i.e., strips) rather than into the deeper tissues (i.e., quarters) of nonintact steaks cut from tenderized subprimals, cooking to internal tempera-

1410 LUCHANSKY ET AL. J. Food Prot., Vol. 72, No. 7 tures of 48.9, 54.4, and 60 C resulted in a average total reduction of E. coli O157:H7 per steak of 2.85, 3.36, and 3.48 log CFU/g, respectively. Studies are in progress to evaluate the fate of non-o157 Shiga-toxin producing isolates in comparison to that of E. coli O157:H7 strains following both blade tenderization and chemical injection and subsequent cooking of nonintact steaks on a commercial gas grill. Collectively, these data, combined with the existing data showing that E. coli O157:H7 contamination of beef subprimals is very infrequent and at low levels, indicate that the risk of illness due to consumption of bladetenderized steaks is extremely small though theoretically possible. This information will be helpful for regulators, restaurateurs, and consumers as they develop and/or practice safe cooking guidelines. ACKNOWLEDGMENTS This project was funded by The Beef Checkoff. We are sincerely grateful to Ellen Sanders, Julie Wade, Boriana Hristova, Jean Smith, Mohammad Koohmaraie, John Cherry, Rosemary Martinjuk, Nellie Osoria, Peggy Williamson, Lihan Huang, and Chris Sommers (USDA/ARS); Randy Huffman (American Meat Institute); Nick Nickelson (Standard Meat Company); Rod Bowling (Smithfield Beef Group); Tim Mohr (USDA/FSIS); Alan Oser (Oser Technologies); Tim Freier (Cargill); and ABC Research Corporation for their excellent insight, technical assistance, and/or valuable resources, all of which were essential for the successful completion of this study. REFERENCES 1. Anonymous. 2005. Beef industry addresses the safety of non-intact beef products. Summary report of the Non-Intact Products Processing Workshop. Coordinated by the National Cattlemen s Beef Association and sponsored by the Cattlemen s Beef Board, Denver, CO. 2. Anonymous. 2006. Best practices for pathogen control during tenderizing/enhancing of whole muscle cuts. Beef Industry Food Safety Council, Denver, CO. 3. California Food Emergency Response Team. 2007. Investigation of an Escherichia coli O157:H7 outbreak associated with Dole prepackaged spinach. Available at: http://www.dhs.ca.gov/ps/fdb/local/ PDF/2006%Spinach%20Final%20redacted.PDF. Accessed 8 December 2008. 4. Centers for Disease Control and Prevention. 2006. Ongoing multistate outbreak of Escherichia coli serotype O157:H7 infections associated with consumption of fresh spinach United States, September 2006. Morb. Mortal. Wkly. Rep. 55:1045 1046. 5. Cox, R. J., J. M. Thompson, C. M. Cunial, S. Winter, and A. J. Gordon. 1997. The effect of degree of doneness of beef steaks on consumer acceptability of meals in restaurants. Meat Sci. 45:75 85. 6. D Sa, E. M., M. A. Harrison, S. E. Williams, and M. H. Broccoli. 2000. Effectiveness of two cooking systems in destroying Escherichia coli O157:H7 and Listeria monocytogenes in ground beef patties. J. Food Prot. 63:894 899. 7. Duffy, G., E. Cummins, P. Nally, S. O Brien, and F. Butler. 2006. A review of quantitative microbial risk assessment in the management of Escherichia coli O157:H7 on beef. Meat Sci. 74:76 88. 8. Gill, C. O., and J. C. McGinnis. 2004. Microbiological conditions of mechanically tenderized beef cuts prepared at four retail stores. Int. J. Food Microbiol. 95:95 102. 9. Gill, C. O., J. C. McGinnis, K. Rahn, D. Young, N. Lee, and S. Barbut. 2005. Microbiological condition of beef mechanically tenderized at a packing plant. Meat Sci. 69:811 816. 10. Grant, J., A. M. Wendelboe, A. Wendel, B. Jepson, P. Torres, C. Smelser, and R. T. Rolfs. 2008. Spinach-associated Escherichia coli O157:H7 outbreak, Utah and New Mexico, 2006. Emerg. Infect. Dis. 14:1633 1636. 11. Jackson, C. J., M. D. Hardin, and G. R. Acuff. 1996. Heat resistance of Escherichia coli O157:H7 in a nutrient medium and in ground beef patties as influenced by storage and holding temperatures. J. Food Prot. 59:230 237. 12. Juneja, V. K. 2002. Thermal inactivation, chap. 2, p. 13 53. In V. K. Juneja and J. N. Sofos (ed.), Control of foodborne microorganisms. Marcel Dekker, Inc., New York. 13. Juneja, V. K., O. P. Snyder, A. C. Williams, and B. S. Marmer. 1997. Thermal destruction of Escherichia coli O157:H7 in hamburger. J. Food Prot. 60:1163 1166. 14. Kennedy, J. E., S. K. Williams, T. Brown, and P. Minerich. 2006. Prevalence of Escherichia coli O157:H7 and indicator organisms on the surface of intact subprimal beef cuts prior to further processing. J. Food Prot. 69:1514 1517. 15. Laine, E. S., J. M. Scheftel, D. J. Boxrud, K. J. Vought, R. N. Danila, K. M. Elfering, and K. E. Smith. 2005. Outbreak of Escherichia coli O157:H7 infections associated with nonintact blade-tenderized frozen steaks sold by door-to-door vendors. J. Food Prot. 68:1198 1202. 16. Lambert, D. L., R. K. Phebus, H. Thippareddi, J. L. Marsden, and C. L. Kastner. 2001. Salmonella risk assessment for production and cooking of non-intact pork products, abstr. PO54, p. 60. Abstr. Annu. Meet. Int. Assoc. Food Prot. International Association for Food Protection, Des Moines, IA. 17. Lorca, T. A., M. D. Pierson, J. R. Claus, J. D. Eifert, J. E. Marcy, and S. S. Sumner. 2002. Penetration of surface-inoculated bacteria as a result of hydrodynamic shock wave treatment of beef steaks. J. Food Prot. 65:616 620. 18. Luchak, G. L., R. K. Miller, K. E. Belk, D. S. Hale, S. A. Michaelsen, D. D. Johnson, R. L. West, F. W. Leak, H. R. Cross, and J. W. Savell. 1998. Determination of sensory, chemical and cooking characteristics of retail beef cuts differing in intramuscular and external fat. Meat Sci. 50:55 72. 19. Luchansky, J. B., J. E. Call, R. K. Phebus, and T. Harshavardan. 2008. Translocation of surface inoculated Escherichia coli O157:H7 into beef subprimals following blade tenderization. J. Food Prot. 71: 2190 2197. 20. Luchansky, J. B., A. Porto-Fett, B. Shoyer, and J. E. Call. 2008. Thermal inactivation of Escherichia coli O157:H7 in blade tenderized beef steaks cooked on a commercial open-flame gas grill, abstr. PV-15, p. 383. Abstr. 21st Int. Committee Food Microbiol. Hyg. (ICFMH) Symposium, Aberdeen, Scotland. 21. Mukherjee, A., Y. Yoon, K. E. Belk, J. A. Scanga, G. C. Smith, and J. N. Sofos. 2007. Cooking destruction of Escherichia coli O157:H7 in beef with marination and tenderization ingredients. Available at: http://www.ansci.colostate.edu/files/research reports/ 07ResearchReports/Avik Cooking destruction of Ecoli in Beef.pdf. Accessed 15 December 2008. 22. National Advisory Committee on Microbiological Criteria for Foods. 2002. Final report: E. coli O157:H7 in blade tenderized beef. Available at: http://www.fsis.usda.gov/ophs/nacmcf/2002/rep blade1. htm. Accessed 15 December 2008. 23. Pflug, I. J., and G. W. Grahame. 2000. Heat treatment, p. 36 60. In B. Lund, T. C. Baird-Parker, and G. W. Gould (ed.), The microbiological safety and quality of food. Aspen Publishers, Inc., Gaithersburg, MD. 24. Phebus, R. K., H. Thippareddi, S. Sporing, J. L. Marsden, and C. L. Kastner. 2000. Escherichia coli O157:H7 risk assessment for bladetenderized beef steaks. Cattlemen s Day 2000. Report of progress 850, p. 117 118. Kansas State University, Manhattan. 25. Rangel, J. M., P. H. Sparling, C. Crowe, P. M. Griffin, and D. L. Swerdlow. 2005. Epidemiology of Escherichia coli O157:H7 outbreaks, United States, 1982 2002. Emerg. Infect. Dis. 11:603 609. 26. Rhee, M. S., S. Y. Lee, V. N. Hillers, S. M. McCurdy, and D. H. Kang. 2003. Evaluation of consumer-style cooking methods for reduction of Escherichia coli O157:H7 in ground beef. J. Food Prot. 66:1030 1034. 27. Schmidt, T. B., M. P. Keene, and C. L. Lorenzen. 2002. Improving consumer satisfaction of beef through the use of thermometers and consumer education by wait staff. J. Food Sci. 67:3190 3193. 28. Sofos, J. N., I. Geornaras, K. E. Belk, and G. C. Smith. 2008. Nonintact whole muscle food safety: the problem and research needs, p. 1 7. Proceedings of the 61st American Meat Science Association

J. Food Prot., Vol. 72, No. 7 FATE OF E. COLI O157:H7 IN COOKED NONINTACT BEEF 1411 Reciprocal Meat Conference. American Meat Science Association, Savoy, IL. 29. Sporing, S. B. 1999. Escherichia coli O157:H7 risk assessment for production and cooking of blade-tenderized beef steaks. Available at: http://www.fsis.usda.gov/oppde/rdad/frpubs/01-013n/blade3. pdf. Accessed 13 November 2008. 30. U.S. Department of Agriculture. 1999. Beef products contaminated with Escherichia coli O157:H7. Fed. Regist. 64:6881 6886. 31. U.S. Department of Agriculture. 2005. HACCP plan reassessment for mechanically tenderized beef products. Fed. Regist. 70:30331 30334. 32. Warren, W. 2002. Characterization of E. coli O157:H7 Subprimal beef cuts prior to mechanical tenderization. National Cattlemen s Beef Association Research Project Summary. Available at: http:// www.beef.org/udocs/e.%20coli%20mech%20tenderization Warren 6 6 03.pdf. Accessed 4 December 2008.