A total of 1,729 E. coli isolates from human and animal samples obtained from different US states were used in this study. Isolates were obtained by the American Type Culture Collection (ATCC) (Manassas, VA, USA) from the E. coli Reference Center (ECRC) at Pennsylvania State University (University Park, PA, USA) and the Centers for Disease Control and Prevention (CDC) (Atlanta, GA, USA) under contract with the US Food and Drug Administration Center for Veterinary Medicine (Rockville, MD, USA). These isolates were recovered from human and animal specimens (e.g., feces, blood, kidney, lymph nodes, urine, cerebrospinal fluid, peritoneal fluid, pleural fluid) submitted to ECRC and CDC from state public health and veterinary diagnostic laboratories. For human isolates obtained from CDC, most acquired during 1948 through the late 1980s were maintained on trypticase soy agar stabs sealed with paraffin and stored at room temperature. Starting in the late 1980s, strains were frozen in trypticase soy broth containing 30% glycerol at −70°C. Similarly, isolates were stored, according to the ECRC standard protocol at −70°C to −80°C in trypticase soy broth containing 30% glycerol until further processing.

Of 1,729 E. coli isolates, 983 (56.8%) were recovered from humans during 1950–2001, and the remaining 746 (43.2%) were recovered from animals during 1962–2002. Three hundred twenty-three (43.2%) E. coli isolates of animal origin were from cattle, 138 (18.5%) from chickens, and 285 (38.2%) from pigs. Fifty percent were from 10 states: Pennsylvania (183, 11.2%), California (100, 5.8%), Ohio (90, 5.2%), Maryland (83, 4.8%), Minnesota (82, 4.7%), Illinois (72, 4.2%), Texas (72, 4.2%), New York (70, 4.1%), North Carolina (61, 3.5%), and Virginia (58, 3.4%). Distribution of isolates by source and year are shown in Table 1.

Each isolate was streaked on trypticase soy agar supplemented with 5% defibrinated sheep blood (Becton Dickinson, Sparks, MD, USA) before antimicrobial drug susceptibility testing. MICs were determined by using the Sensititer automated antimicrobial susceptibility system (Trek Diagnostic Systems, Cleveland, OH, USA) according to the manufacturer’s instructions. Results were interpreted according to National Committee for Clinical and Laboratory Standards criteria (16) where available (Table 2). Antimicrobial drugs tested were ampicillin, amoxicillin/clavulanic acid, cefoxitin, ceftiofur, cephalothin, ceftriaxone, ciprofloxacin, nalidixic acid, streptomycin, gentamicin, kanamycin, chloramphenicol, tetracycline, sulfonamide, and trimethoprim/sulfamethoxazole. E. coli ATCC 25922 and ATCC 35218, Enterococcus faecalis ATCC 29212, Staphylococcus aureus ATCC 29213, and Pseudomonas aeruginosa ATCC 27853 were used as quality control organisms in MIC determinations. Multidrug resistance was defined as resistance to >3 classes of antimicrobial drugs.

The Mann-Kendall test, a nonparametric statistical test, was performed to detect a monotone increasing or decreasing resistance trend over time. Magnitude of annual change was estimated by using a slope parameter, Q, and the Sen nonparametric method (20). Calculations were performed by using the Excel (Microsoft, Redmond, WA, USA) template Mann-Kendall test for trend and Sen slope estimates (21). For time series <10 annual percentage resistance values, significance of the trend was determined from the exact distribution of the S test statistic, and a normal approximation (z statistic) was used when there were >10 values. Significance was assessed at 4 levels (α = 0.001, 0.01, 0.05, and 0.1); p values <0.05 were considered significant. Comparisons of drug resistance profiles between different sources (human, cattle, chicken, and pigs) were conducted by using the χ² test; p values <0.05 was considered significant.

Overall, 934 (54.0%) of 1,729 E. coli were resistant to >1 antimicrobial drug. As expected, the most common resistance phenotypes were to older drugs such as tetracycline (40.9%) (introduced in 1948), sulfonamide (36.2%) (introduced in 1936), streptomycin (34.2%) (introduced in 1943), and ampicillin (24.1%) (introduced in 1961). A much smaller number of isolates were resistant to antimicrobial drugs introduced for clinical use since 1980, such as amoxicillin/clavulanic acid (5.6%) (introduced in 1984), ceftriaxone (2.4%) (introduced in 1984), ceftiofur (2.3%) (introduced in 1988), and ciprofloxacin (0.4%) (introduced in 1987) (Table 2).

When analyzed by source, E. coli isolates of animal origin were more resistant than those of human origin. Among 983 human isolates, resistance was observed most often to sulfonamide (19.9%), followed by tetracycline (18%) and ampicillin (16.5%). No human E. coli isolates showed resistance to ciprofloxacin, and 1 isolate (0.1%) from 1997 showed resistance to ceftiofur, ceftriaxone, and gentamicin. Of 746 isolates recovered from animal sources, 531 (71.1%) were resistant to tetracycline, 441 (59%) to streptomycin, 431 (57.7%) to sulfonamide, 277 (37.1%) to kanamycin, and 255 (34.1%) to ampicillin. Among animal E. coli isolates, the rate of resistance was significantly higher in cattle isolates than in pig isolates for chloramphenicol (p = 0.039), amoxicillin/clavulanic acid (p = 0.03), sulfonamide (p = 0.038), and trimethoprim/sulfamethoxazole (p = 0.022). There was a significant difference in resistance rate between cattle and chicken isolates to ceftiofur (p = 0.008), ceftriaxone (p = 0.008), chloramphenicol (p = 0.011), and kanamycin (p = 0.037) (Table 2).

Antimicrobial drug resistance was observed for drugs tested at different frequencies (Table 2). Seven hundred ninety-six (46%) E. coli isolates analyzed were susceptible to all 15 drugs tested. Among these pan-susceptible isolates, 637 (80%) were from humans, 69 (8.7%) from cattle, 60 (7.5%) from pigs, and 29 (3.6%) from chickens. Approximately 65% of human isolates were pan susceptible, compared with ≈20% of cattle, chicken, and pig isolates (Table 3). The proportion of pan-susceptible E. coli isolates decreased from 73.9% in 1950–1959 to 18.5% in 2000–2002. Conversely, multidrug resistance increased from 7.2% in 1950–1959 to 63.6% in 2000–2002 (Figure 1). Five hundred seventy (32.9%) E. coli isolates showed multidrug-resistant phenotypes, and 176 (10.2%) showed resistance to >5 drug classes. A larger proportion of multidrug-resistant isolates was recovered from animals than humans (Figure 2). One hundred ninety-one (59.1%) isolates from cattle, 153 (53.7%) from pigs, and 76 (55.1%) from chickens were resistant to >3 drug classes (Table 3), compared with 15.3% from humans. Two strains showed resistance to all 15 drugs tested; both strains were recovered from cattle in 2001.

Concurrent resistance to tetracycline and streptomycin was the most common co-resistance phenotype (29.7%), followed by resistance to tetracycline and sulfonamide (29.0%); tetracycline, sulfonamide, and streptomycin (23.9%); tetracycline and ampicillin (18.8%); and tetracycline, ampicillin, streptomycin, and sulfonamide (12.9%). A total of 130 (92.9%) of 140 chloramphenicol-resistant E. coli isolates were also resistant to tetracycline. Resistance to ceftriaxone, ceftiofur, and ciprofloxacin was rare and found only in isolates resistant to ≥7 drugs. More than 80% of these isolates were resistant to ampicillin, amoxicillin/clavulanic acid, cephalothin, cefoxitin, streptomycin, tetracycline, and sulfonamide.

The major goal of this study was to document antimicrobial drug resistance among historical bacteria from humans and animals to associate emergence of resistance with approval of new antimicrobial classes. Animal E. coli isolates showed an increasing resistance trend to 11 antimicrobial agents (ampicillin, sulfonamide, tetracycline, cephalothin, trimethoprim/sulfamethoxazole, streptomycin, chloramphenicol, cefoxitin, gentamicin, amoxicillin/clavulanic acid, and kanamycin), and human E. coli isolates showed an increasing trend in resistance only to ampicillin, sulfonamide, and tetracycline (Figure 3, Table 4).

Human E. coli isolates showed an increased resistance trend for ampicillin (0.59%/year, 95% CI 0.38%–0.81%; p<0.001), sulfonamide (0.49%/year, 95% CI 0.23%–0.73%; p<0.001), and tetracycline (0.45%/year, 95% CI 0.22%–0.70%; p<0.001), and this trend ranged during the study period from 0% to 66.7% for ampicillin, 0% to 50% for sulfonamide, and 0% to 58% for tetracycline (Table 4). The resistance rate to ampicillin in animal E. coli isolates was similar to that in human isolates and ranged from 0% to 69.4%. In contrast, resistance rates for sulfonamide and tetracycline among animal E. coli isolates ranged from 0% to 73.7% and 0% to 85.5%, respectively, and were higher than those for human isolates. There was a linear increasing trend in resistance to ampicillin (0.97%/year, 95% CI 0.60%–1.33%; p<0.001), sulfonamide (1.11%/year, 95% CI 0.34%–1.85%; p<0.01), and tetracycline (0.7%/year, 95% CI 0.21%–1.21%; p<0.01) (Table 4).

Cephalothin resistance significantly increased over time (0.43%/year, 95% CI 0.08%–0.77%; p<0.05) among animal E. coli isolates. In contrast, a decreasing resistance trend to cephalothin (−0.2%/year, 95% CI −0.34% to 0%; p<0.05) was observed among human isolates (Table 4). Although animal E. coli isolates showed an increasing rate of streptomycin resistance over time (0.9%/year, 95% CI 0.14%–1.41%; p<0.05), there was no significant increase among human E. coli isolates during the study period (0.15%/year, 95% CI −0.10% to 0.39%; p>0.1) (Table 4). In our study isolates, gentamicin resistance was observed in the 1980s among animal E. coli isolates but not until the late 1990s in human isolates. In animal E. coli isolates, the prevalence of gentamicin resistance increased from 0% during the 1970s to 28.1% during 2000–2002, and an increasing trend for resistance to gentamicin (1.28%/year, 95% CI 0.90%–1.69%; p<0.001) was first observed in the 1980s and reached a prevalence of 40% in 2002. Chloramphenicol resistance varied widely between human and animal isolates (yearly range 0%–46.8% among animal isolates and 0%–20% among human isolates). A linear increase in chloramphenicol resistance was observed among animal isolates (p<0.01), which increased 0.30%/year. Ceftiofur-resistant and ceftriaxone-resistant strains were not detected until 1990–1999 among human and animal E. coli isolates. None of the human E. coli isolates showed resistance to ciprofloxacin.

There was no monotonic resistance trend for trimethoprim/sulfamethoxazole, chloramphenicol, cefoxitin, ceftiofur, ceftriaxone, gentamicin, amoxicillin/clavulanic acid, nalidixic acid, and kanamycin among human E. coli isolates. Similarly, animal E. coli isolates did not show a monotonic resistance trend for ceftiofur, ceftriaxone, ciprofloxacin, and nalidixic acid.