Optimal regimens based on PK/PD cutoff evaluation of ceftiofur against Actinobacillus pleuropneumoniae in swine

Background Actinobacillus pleuropneumoniae formerly known as Haemophilus pleuropneumoniae, can cause pleuropneumoniae in pigs, which lead to significant mortality. Ceftiofur was the first cephalosporin antibiotic used in animals, which was effective against gram-negative and gram-positive bacterium. This study aimed to formulate a rational dosage strategy and review the preceding recommended dosage based on PK/PD modeling and Establish Clinical breakpoint of ceftiofur against Actinobacillus pleuropneumoniae based on the pharmacodynamic-pharmacokinetic cutoff. Results The epidemiologic cutoff value was 0.125 μg/mL. The results of the pharmacodynamic study showed that the MICs of BW39 were 0.5 μg/mL and 1 μg/mL in vitro and ex-vivo, respectively. The minimal bactericidal concentrations (MBCs) under in vitro and ex vivo conditions were both 1 μg/mL. The time-killing profiles of ceftiofur against BW39 were time-dependent with a partly concentration-dependent pattern. Based on the inhibitory sigmoid Emax model, the AUC24 h/MIC values for the bacteriostatic, bactericidal, and elimination effects in serum were 45.73, 63.83, and 69.04 h for healthy pigs separately. According to the Monte Carlo simulation, the COPD was calculated as 2 μg/mL, and the optimized dosage regimen of ceftiofur against Actinobacillus pleuropneumoniae to achieve bacteriostatic, bactericidal, and elimination effects over 24 h was 2.13, 2.97, and 3.42 mg/kg for the 50% target attainment rate (TAR) and 2.47, 3.21, and 3.70 mg/kg for the 90% TAR respectively. Conclusions In conclusion, we reveal the EOFF and PK/PD cutoff values of ceftiofur against A. pleuropneumoniae in piglets. However, with the paucity of clinical data for ceftiofur to establish a clinical cutoff against A. pleuropneumoniae, the PK/PD cutoff value of 2 μg/mL will be recommended as surrogate. According to the PK/PD data and the MIC distribution in China, the single bactericidal dose was 3.21 mg/kg for the 90% target, which would be more able to cure Actinobacillus pleuropneumoniae and avoid the emergence of resistance for clinical ceftiofur use in piglet.


Background
Porcine pleuropneumonia caused by Actinobacillus pleuropneumoniae is a highly contagious respiratory disease that causes hemorrhage, purulent and fibrous pleuropneumonia. The disease is widely distributed and bring severe losses to the pig industry. The morbidity of the disease can be as high as 100%, but usually varies from 30 to 50% [1]. There are currently 18 serotypes of Actinobacillus pleuropneumoniae [2,3], and the prevalent serotypes vary in different countries and regions, which may lead to piglet infected with multiple serotypes.
Ceftiofur was the first third-generation cephalosporin antibiotic used in animals. It is effective for gramnegative bacteria and gram-positive bacteria and has a good clinical therapeutic effect against animal respiratory diseases [4]. The FDA has approved ceftiofur hydrochloride, ceftiofur sodium and ceftiofur crystal-free acid for the treatment of respiratory diseases of pigs, cattle, horses, goats and sheep [5][6][7].
Clinical breakpoint are MIC values, which are used in clinical microbiology laboratories to categorize microorganisms as clinically [8]. Establishing clinical breakpoint contains three parts: (i) an epidemiological cut-off value (ECOFF), described by EUCAST as the ECOFF, it defines the upper end(> 95%) of the wild type MIC distribution. (ii) PK/PD breakpoint set by EUCAST is generally taken as highest MIC for which a selected PK/ PD index can be achieve in the target population, given the standard regimes and taking into account the lower 95-99% prediction intervals for the population. (iii) clinical cut-off described by EUCAST in human medicine, it is related to clinical outcomes(cure vs non-cure) and requires specific investigations during prospective clinical trials [9]. Although the CLSI has established the breakpoint of ceftiofur against Actinobacillus pleuropneumoniae, considering that the use of antibiotics and drug resistance in China is totally different, it is necessary to formulate the relevant criteria for ceftiofur against Actinobacillus pleuropneumoniae in China based on the local situation.
The abuse of antibiotics and unreasonable doses of antimicrobial agents are the main factors, which contribute to the development of resistance [10,11]. Existing data have suggested that the resistance rate of Actinobacillus pleuropneumoniae to tetracyclines and penicillin reached 90% and showed different degrees of resistance to macrolides and aminoglycosides. However, Actinobacillus pleuropneumoniae is still highly sensitive to ceftiofur [12][13][14] and Clinical breakpoint should be set up for ceftiofur against Actinobacillus pleuropneumoniae. Therefore, we need to use scientific methods to determine the reasonable dosage of ceftiofur and establish drug resistance criteria to protect the clinical efficacy of ceftiofur. Previous PK study [15] has shown ceftiofur sodium and ceftiofur hydrochloride has similar therapeutic efficacy. And the recommended dosage of ceftiofur for the treatment of swine respiratory disease in the U.S. and in Europe ranged from 3 to 5 mg of ceftiofur equivalents (CE)/kg body weight. PK/PD integration modeling data can provide an optimal drug dosage strategy, reducing resistance development, which is a key method to evaluate the clinically relevant relationship between time, drug concentration, and effect. However, there are no data supporting the rational dosage of ceftiofur against Actinobacillus pleuropneumoniae based on the PK/PD model in China.
The purposes of this investigation were to (i) study the pharmacokinetic/pharmacodynamic of ceftiofur against Actinobacillus pleuropneumoniae, (ii) formulate a rational dosage strategy and review the preceding recommended dosage based on PK/PD modeling for ceftiofur against Actinobacillus pleuropneumoniae. (iii) establish a clinical breakpoint based on pharmacokinetic/pharmacodynamic cutoff values.

MIC distribution and epidemiologic cutoff value
The MIC distribution of ceftiofur against Actinobacillus pleuropneumoniae is shown in Fig. 1. The MIC values ranged from 0.0075 to 4 μg/mL. The MIC 50 and MIC 90 were calculated to be 0.015 and 0.5 μg/mL. We selected the serotype 1 strain App BW39 whose MIC was similar to MIC90(0.5 μg/mL) to analyzed for PD study in broth and serum. The ECV was 0.125 μg/mL, encompassing 99.9% of the wild-type isolates.

MICs, MBCs, MPC and PAE
The MIC values of ceftiofur against App BW39 were 0.5 μg/mL and 1 μg/mL in vitro and ex vivo, and the MBC values were both 2 μg/mL in vitro and ex vivo. The MBC/MIC ratios were 2 and 1 in vitro and ex vivo indicating that ceftiofur is a bactericidal drug. The MPC value of ceftiofur against App BW39 was 1.6 μg/mL, and the MSW was 1-1.6 μg/mL. The post antibiotic effect (PAE) values after exposure to the concentrations of 1MIC, 2MIC and 4MIC were 0.33, 0.41 and 0.66 h for 1 h exposure and 0.73, 0.92 and 1.17 h for 2 h exposure, respectively ( Table 1). The result suggested PAE had a positive correlation with exposure time. PAE was also increased with drug concentration, but the degree of enhancement was weak.

Time-killing curves
The time-killing curves of ceftiofur against BW39 in vitro and ex vivo were shown in Figs. 2 and 3. When the drug concentration was 1/2 MIC-1 MIC, bacterial growth was not inhibited. Concentrations between 1 MIC and 2 MIC showed a slight inhibitory effect on the bacterial growth. Ceftiofur achieved a maximum bactericidal effect from 4 MIC-8 MIC. The ex vivo killing curve showed that the plasma collected between 0.33 and 2 h achieved the maximum bactericidal effect at the highest ceftiofur concentration. When the concentration was below the MIC at 96 h, ceftiofur no longer inhibited the growth of bacteria.

Pharmacokinetic analysis of ceftiofur in plasma
LOD and LOQ of HPLC methods for detecting DFC were 0.03 μg/mL and 0.08 μg/mL, respectively. The linear range of the standard and working curve of DFC ranged from 0.1 to 50 μg/mL, with a coefficient of determination (R 2 ) of 0.9999 and 0.9987. The recovery of ceftiofur in plasma ranged from 82.95 to 84.38%. The pharmacokinetic parameters calculated using noncompartment model analysis after I.M. (5 mg/kg BW) administration in healthy pigs are shown in Table 2 (Parameters of compartmental model is shown in Additional file 1: Table1). DFC concentrations were measured at different times shown in Fig. 4.

PK/PD integration and analysis
The PK/PD indices were determined by integrating the in vivo PK data and the ex vivo time-killing curve. The ratios of C max/MIC, AUC 24 Table 3. The values of the AUC24 h/MIC ratio required for bacteriostatic activity (E = 0), bactericidal activity (E = − 3), and bacterial elimination (E = − 4) were 45.73, 63.83, and 69.04 h.

Dose estimation
According to the PK-PD integration model, the PK-PD index, AUC 24 h/MIC, was altered based on different outcomes in healthy swine. The prediction of ceftiofur curing A. pleuropneumoniae diseases for the 50 and 90% targets was calculated by the Monte Carlo simulation and dosage equation, which were shown in Fig. 5. The doses predicted to exhibit bacteriostatic, bactericidal, and elimination effects for A. pleuropneumoniae over 24 h were 2.13, 2.97, and 3.42 mg/kg for the 50% target attainment rate (TAR) and 2.47, 3.21, and 3.70 mg/kg for the 90% TAR, based on Crystal Ball software.

CO PD determination with Monte Carlo simulation
According to the MIC distribution, AUC 24 h derived from healthy pigs and the PK-PD target (AUC 24 h / MIC) ex of 63.83 h, 10,000 subjects were simulated by Monte Carlo simulation. The PTA (possibility target achievement) values were calculated as 88.94, 100 and 100% when the MICs (PK-PD cut-off) were defined as 4, 2 and 1 μg/ml, respectively (Table 4). Therefore, the CO PD for ceftiofur against Actinobacillus pleuropneumoniae was defined as 2 μg/mL, for the current dose of 5 mg/kg by IM route.

Discussion
Ceftiofur, whether administered as hydrochloride or sodium salt, was metabolized rapidly to desfuroylceftiofur. The plasma half-life of ceftiofur sodium after intravenous dosing in swine was approximately 10 min [16]. The previous study [17] has tested the susceptibility of ceftiofur and its metabolite desfuroylceftiofur against Actinobacillus pleuropneumoniae(n = 50), the MIC90 of ceftiofur was 0.0078 μg/mL, and 0.015 μg/mL for desfuroylceftiofur, which suggested ceftiofur can be predictive of in vitro activity of ceftiofur and its major metabolite, desfuroylceftiofur. In our study, we determined the vitro susceptibility of ceftiofur, instead of desfuroylceftiofur, against 135 Actinobacillus pleuropneumoniae strains.
The MIC values ranged from 0.0075 μg/mL to 4 μg/mL, and the MIC 50 and MIC 90 values were 0.015 μg/mL and 0.5 μg/mL, respectively, which were slightly different from the results of ceftiofur against 50 Actinobacillus pleuropneumoniae strains (MIC ranged from 0.0039-0.0015 μg/mL) [17]. The epidemiologic cutoff was established based on the pretest of drug susceptibility and then analyzed by statistical methods. At present, CLSI and EUCAST guidelines mainly involve analysis by nonlinear regression [18]. EUCAST based on the nonlinear regression complied the ECOFFinder software [19]. In this study, the epidemiologic cutoff was 0.125 μg/L according to the ECOF-Finder results.
There are currently 18 serotypes of Actinobacillus pleuropneumoniae [2,3]. The prevalent serotypes are different all over the world. The main prevalent serotypes in Canada are serotypes 1, 3 and 5, those in Europe are serotypes 1, 2, 5, 7 and 9 [20], and those in the United States are serotypes 1, 5 and 7 [21]. However, serotypes 1, 3,5 and 7 were the main prevalent serotypes in China [22]. Moreover, serotype 1 was more virulent than other serotypes [23,24] Therefore, A strain App BW39, whose MIC was similar to MIC90(0.5 μg/mL) and serotype was serotype 1, was analyzed for PD study in broth and serum. According to the MIC and MBC values in vitro and ex vivo, the MBC/MIC ratio was 2 and 1. Therefore, ceftiofur is a bactericidal drug with an MBC/MIC ratio < 4 [25]. The results of in vitro and ex vivo killing curves showed, when the drug concentration was 1/2 MIC-1 MIC, bacterial growth was not inhibited. With the concentration increasing, the bactericidal effect increased and achieved the maximum bactericidal effect at 4MIC-8MIC. These findings suggest that ceftiofur has a time-dependent inhibition activity with a partly concentration-dependent pattern against Actinobacillus pleuropneumoniae both in vitro and ex vivo [26].
To establish a pharmacokinetic/pharmacodynamics cutoff, we needed to obtain the PK-PD parameters, MIC distribution and pharmacodynamic targets. The most appropriate PK/PD index to determine the antibacterial efficacy and predict the therapeutic efficacy for β-lactam antibiotics is ƒ%T > MIC [9]. The killing curve showed  that ceftiofur had a time-dependent inhibition activity same with a partly concentration-dependent pattern against Actinobacillus pleuropneumoniae. Moreover, PK/ PD index depends on the shape of the plasma exposure curve which may differ widely between the many modalities of Antimicrobial drug administration, such as antimicrobial drug was incorporated in food or drinking water or administrated by oral. VetCAST suggested using AUC/MIC as a default index. The index of the AUC/MIC ratio was also used to describe the character of the antibacterial activity of time-dependent killing with prolonged persistent efficacy [27]. In our manuscript, the PK/PD index (AUC/MIC), using an inhibitory sigmoidal Emax model, showed a favorable correlation (r 2 = 0.9967) with the predicted antibacterial efficacy. For these reasons, the AUC/MIC parameter was regarded as the most appropriate PK/PD index to describe the antibacterial activity of ceftiofur in serum. An AUC/MIC ratio > 125 h is generally considered the best activity indicator [28,29]. However, the target parameters might be different for various drugs, target organisms or bacteria [30].  [3]. The PK of ceftiofur may differ from that of diseased pigs. Previous study has demonstrated that infected pigs may have higher values of Vz/F and CL/F, but lower values of AUC, MRT and t1/2z compared with non-infected pigs.
PD-PD integration modeling can be used to select rational dose regimes in veterinary medicine [29]. Based on the dose estimation equation and Monte Carlo simulation, the 'fu' in the equation was the free fraction of the drug in plasma. Ceftiofur had a high protein binding rate, which was 90% in cow and adult cattle [32,33],The EMA [34] reported that the protein binding of ceftiofur and its metabolite DFC in swine was 0.7, which was applied in this research. According to the MIC distribution in China, the predicted daily doses for the 50 and 90% targets of ceftiofur with bactericidal activity against Actinobacillus pleuropneumoniae were 2.97 and 3.21 mg/kg body weight, respectively.
Generally, Clinical breakpoint is determined by epidemiological cut-off value, PK/PD breakpoint and Clinical cutoff [24,25]. Under the clinically recommended   hydrochloride for a PTA ≥ 90% was 3.21 mg/kg. Previous study [3] has shown the similarity of the pharmacokinetic parameters of the sodium and hydrochloride formulations of ceftiofur, similar therapeutic efficacy therefore can be inferred for the two products. So, considering the MIC distribution in China, the dose of ceftiofur could be 3.21 mg/kg, However, it is practically difficult to determine a clinical cutoff in veterinary medicine [24,26]. Owing to the paucity of relevant data to bridge the relationship between MIC and clinical cure. In our study, CO PD values contains the ECOFF values, and the CO PD (2 μg/mL) will therefore be recommended as surrogate.

Conclusions
In conclusion, we reveal the EOFF and PK/PD cutoff values of ceftiofur against A. pleuropneumoniae in piglets. However, with the paucity of clinical data for ceftiofur to establish a clinical cutoff against A. pleuropneumoniae, the PK/PD cutoff value of 2 μg/mL will be recommended as surrogate. Moreover, according to the PK/PD data and the MIC distribution in China, the single bactericidal dose was 3.21 mg/kg for the 90% target, which would be more able to cure Actinobacillus pleuropneumoniae and avoid the emergence of resistance for clinical ceftiofur use of ceftiofur in piglet.

Chemicals and reagents
The standard ceftiofur hydrochloride (det. Purity: 97.9%) purchased from Dr. Ehrenstorfer (Augsburg, Germany) and 5% ceftiofur hydrochloride injection purchased from Pfizer Animal Health Co., Ltd., Canada, were used during the study. All the chemical reagents used were HPLC grade.

Determination of antimicrobial susceptibility and epidemiologic cutoff
The susceptibility of 135 Actinobacillus pleuropneumoniae strains to ceftiofur was determined by inoculating the strains on TSA agar plates supplemented with newborn calf serum and 1% NAD and incubating the plates at 37°C in an atmosphere containing 5% CO 2 for 24 h; the susceptibility was measured using the standard agar dilution method with concentrations of ceftiofur between 0.00375-32 μg/ml, according to CLSI [35] protocols. The MIC distribution was constructed and converted into a cumulative log-normal distribution. Then, nonlinear regression was employed to fit the cumulative log 2 -transformed MIC data to obtain a range of optimum wild-type MIC distributions, which contained the wild-type MIC in the range of 0.1 and 99.9%, and to calculate the probability of MIC data falling within the cutoff range. The optimum fit was defined as the fit where the difference between the estimate of the isolate number and the actual number was minimal. The cutoff value would encompass at least 95% of the wild-type isolates [18]. A wild-type cutoff (CO WT ) was developed based on the MIC distribution following CLSI M37-A3. The CO WT value was calculated by ECOFFinder software (J. Turnidge, Kahlmeter, and Kronvall 2006), which is available on the CLSI website (https://clsi.org/education/microbiology/ecoffinder/).

Pharmacodynamics of ceftiofur against BW39 MIC, MBC and MPC determination
The minimum inhibitory concentration (MIC) of ceftiofur against BW39 (serotype 1) was determined by the broth micro dilution method according to the CLSI (Clinical and Laboratory Standards Institute, 2015) and in serum, as vitro MIC and ex vivo MIC. The MIC was determined as the lowest ceftiofur concentration that visibly inhibited the growth of bacteria at the end of the 24 h incubation period. For the MBC of ceftiofur against BW39, 100 μL from each well was subjected to 10-fold or more dilution with broth and serum, as vitro MBC and ex vivo MBC; 10 μL of each solution was spread on TSA agar plates and incubated at 37°C for 24 h for colony forming unit (CFU) counting. The MBC was defined as the lowest drug concentration that resulted in a 99.9% reduction in the bacterial density. The mean was expressed as the final result. The mutant prevention concentration (MPC) was determined by the agar method according to the procedures of Blondeau [36]. Exponential growth phase bacteria were pelleted by centrifuging at 3000 rpm at 4°C. The pellet was then diluted to 3 × 10 10

Bacterial growth and time-killing curve of BW39 in vitro and ex vivo
The BW39 isolate was selected to determine the growth curve and time-killing curve in TSB broth and serum. The in vitro and ex vivo growth curves of the BW39 isolate were established by plotting time versus log 10 CFU/ mL. An aliquot of 5 mL of BW39 grown to mid-log phase with a starting inoculum of 10 6 CFU/mL was added to 5 mL of TSB broth supplemented with serial concentrations of ceftiofur corresponding to 0 × MIC, production. One hundred microliter cultured medium was mixed with 0.9 mL TSB medium and cultured 24 h at 37°C with 5% CO 2 . One hundred microliter samples were taken at 0, 1, 2, 4, 6, 8, and 12 h, and serially diluted 10-fold with sterile physiological saline to count cells. Each treatment was performed four times. Growth curve for A.pleuropneumoniae BW39, at different ceftiofur concentrations were established, and T (time required for bacterial numbers to be 10 times higher than 0 h in the test groups) and C (time required for the bacterial numbers to be 10 times higher than 0 h in control groups) values were calculated. PAE was calculated as the difference between T and C (PAE = T -C) [26].

Pharmacokinetics of ceftiofur in plasma of pigs Animals
Six 6-week-old healthy castrated crossbred piglets (Duroc×Landrace×Yorkshire) pigs with an average weight of 15 ± 2 kg were purchased from the Livestock and Poultry Breeding Center of Hubei Province (Wuhan, China). The animals were acclimatized for a period of 1 week before the experiment. The temperature and relative humidity of the housing environment were kept at 18-25°C and 45-65%, respectively. All animal experiment procedures were approved by the Institutional Animal Care and Use Committee at Huazhong Agricultural University (HZAUSW-2016-007).

Samples collection
Six pigs were received ceftiofur hydrochloride injection at a dose of 5 mg/kg·b.w by intramuscular injection of neck. Blood samples (5 mL) from each pig of each group were gently collected from the jugular vein at 0, 0.33, 0.66, 1, 1.5, 2, 3, 5, 8, 12, 24, 36, 48, 72 and 96 h. Plasma samples were obtained by centrifuging the blood at 3500 rpm/min for 10 min, and the samples were stored at − 20°C prior to the analysis and analysis within 3 days after sampling.

Sample analysis
DFC concentration were measured by HPLC described by previous study [26] to represent ceftiofur plasma concentration, as ceftiofur is rapidly metabolized to DFC in piglets. Extraction: 0.5 mL plasma was mixed with 7 mL of 0.4% DTE-borate buffer. The mixture was incubated for 15 min at 50°C in a water bath, with a 10 s vortexing every 3 min. Samples were then centrifuged after cooling to 25°C, then the supernatant was collected.
Solid phase extraction: An Agilent HLB column (60 mg/3 cc) was activated and equilibrated consecutively with 3 mL methanol and ultrapure water. Extracted materials were added to the HLB column and a flow rate set at 1 mL/min. The column was then eluted with 5 mL methanol, after which the eluate was concentrated by nitrogen at 35°C. The concentrated solution was vortexed with 0.5 mL ultrapure water, sonicated for 5 min. The DFC standard was added to 0.5 mL plasma (to achieve final concentrations of 0.1, 0.25, 0.5,1.0, 2.5, 5.0, 10.0, 20.0 and 50 μg/mL) and prepared with same process as samples from test groups.
The quantitative analysis of ceftiofur in plasma was performed with a Water 2695 series HPLC with a UV detector at a wavelength of 266 nm. A ZORBAX Stable Bond-C18 column (250 mm × 4.6 mm, i.d. 5 μm, Agilent) was used to achieve chromatographic separation. The mobile phase consisted of 0.1% trifluoroacetic acid (phase A) and acetonitrile (phase B) at a flow rate of 1 mL/min at 30°C with isometric elution conditions (86, 14, v/v).
The concentration-time data for ceftiofur in plasma samples harvested from healthy pigs were analyzed by WinNonlin 5.2.1 software (Pharsight Corporation, Mountain View, CA, USA) to obtain the pharmacokinetic parameters.

Pharmacokinetic/pharmacodynamic integration and dose estimations
There are three standard indices (ƒ%T > MIC, ƒC max / MIC, and ƒAUC 0-24 h /MIC) for an antibiotic [37]. The inhibitory sigmoid Emax model (Hill equation) was analyzed the integration of AUC24h/MIC ratio in vitro and bacteria count change (CFU/ml) in serum during 24 h incubation. The model equation was described as follows: where E is the antibacterial effect measured as the change in the bacterial count (log 10 CFU/mL) in plasma sample after 24 h of incubation compared with the initial incubation, E 0 is the change in log 10 difference in bacterial count in the control sample after 24 h of incubation; E max is the maximum antibacterial effect determined as the difference in log 10 CFU/mL in the sample after the incubation, EC 50 is the PK/PD parameter value producing 50% of the maximum antibacterial effect; C is the PK/PD parameter value in the effect compartment (the ex vivo site, that is plasma); and N is the Hill coefficient, which describes the steepness of the PK/PD parametereffect curve.
PK/PD parameter values corresponding to the E value (derived from the sigmoid E max equation) in plasma were used to deduce an optimal dose regimen. The potential optimal dosage was calculated using the following equation: where MIC is the ex vivo minimum inhibitory concentration; (AUC 24 h /MIC) ex is the target end point for optimal efficacy; CL is the clearance; fu is the free fraction of ceftiofur in plasma; and F is the bioavailability of ceftiofur. The distribution probabilities for predicted daily dosage were performed to achieve simulated 50 and 90% TAR under 1000 trials with Crystal Ball software (version 11.1.2, Oracle, United States).

Monte Carlo simulation and the pharmacokinetic/ pharmacodynamics cutoff (CO PD )
A Monte Carlo simulation (MCS) with 10,000 iterations was conducted using Crystal Ball software (version 7.2.2) (Oracle, United States) based on PK parameters and calculated PK/PD targets (AUC 24h /MIC) when each possible MIC and the target AUC 24h /MIC achieving a bactericidal action (E = − 3). The AUC 24h was assumed to be normally. CO PD was defined as the maximal MIC value at which the corresponding PTA was ≥90% [38].