Here, we present a protocol for evaluating the synergistic activity of fosfomycin in combination with ceftriaxone, meropenem, and gentamicin against carbapenem-resistant Escherichia coli isolates using the agar dilution method. Synergy was observed with meropenem and gentamicin, but not with ceftriaxone.
Review Article
Protocol for Agar Dilution Checkerboard Analysis of Fosfomycin-Based Antibiotic Combinations Against Carbapenem-Resistant Escherichia coli
Opens in a new tab
In This Article
Summary
Loading...
$$\rightleftharpoonup{xx}$$
$$\longleftharp{xx}$$,
$$\longrightharp{xx}$$,
Abstract
Loading...
$$\rightleftharpoonup{xx}$$
$$\longleftharp{xx}$$,
$$\longrightharp{xx}$$,
Due to the limited treatment options for carbapenem-resistant Enterobacterales, combination therapy regimens have gained increasing importance. This study aimed to investigate fosfomycin resistance rates and evaluate the in vitro synergistic interactions of fosfomycin in combination with meropenem, gentamicin, and ceftriaxone against carbapenem-resistant Escherichia coli isolates. Sixty carbapenem-resistant E. coli strains isolated from various clinical samples (urine, blood, aspirate, wound, cerebrospinal fluid, endotracheal aspirate) obtained from patients in clinical wards and intensive care units of Dicle University Medical Faculty Hospitals were included in the study. Species identification was performed using an automated microbial identification system. Antibiotic susceptibility testing and confirmation of carbapenem resistance were performed by disc diffusion with ertapenem, imipenem, and meropenem. Fosfomycin resistance and interactions among antibiotics were assessed by agar dilution. In addition, antibiotic combinations were evaluated across serial two-fold concentration gradients using the agar dilution-based protocol, and fractional inhibitory concentration indices (FICI) were calculated to determine synergy, indifference, or antagonism. Resistance rates were 76.6% for fosfomycin, 61.6% for meropenem, 85.0% for gentamicin, and 98.3% for ceftriaxone. Synergistic interactions were observed in 35% of isolates for the fosfomycin–meropenem combination, 33.4% for the fosfomycin–gentamicin combination, and 1.7% for the fosfomycin–ceftriaxone combination. Antagonism was detected in only one isolate in the fosfomycin–meropenem combination. Combinations of fosfomycin with meropenem and gentamicin showed in vitro synergistic activity against carbapenem-resistant E. coli. Further in vitro studies are required to confirm these findings.
Introduction
Loading...
$$\rightleftharpoonup{xx}$$
$$\longleftharp{xx}$$,
$$\longrightharp{xx}$$,
The Enterobacterales family comprises a highly diverse group of gram-negative bacilli, many of which are of significant clinical importance1. Among them, Escherichia coli is the most frequently isolated species in clinical microbiology laboratories and is widely associated with infections involving nearly all organ systems, including urinary tract infections, wound infections, pneumonia in immunocompromised patients, gastroenteritis, and neonatal meningitis. It is also a major cause of Gram-negative sepsis and endotoxin-mediated septic shock. Beyond human medicine, E. coli is also an important veterinary pathogen, causing respiratory and systemic infections in various animal species, including equines, extending the host range perspective of E. coli pathogenesis across human and animal hosts2,3,4.
Extended-spectrum beta-lactamase (ESBL)-producing Enterobacterales, particularly E. coli and Klebsiella pneumoniae, have been associated with increased mortality, prolonged hospitalization, and higher healthcare costs. The widespread use of antibiotics is a major driver of multidrug resistance (MDR) among these organisms1. Resistance mechanisms include carbapenemase production, ESBLs, aminoglycoside-modifying enzymes (AMEs), mutations conferring quinolone resistance, and alterations in porin and efflux systems. In E. coli and Klebsiella pneumoniae, resistance is mainly plasmid-mediated, facilitating horizontal gene transfer. Such resistance traits are also increasingly reported in animal-origin E. coli, including cattle isolates, supporting the One Health concept and indicating potential zoonotic transmission risk2,3.
Carbapenems are considered last-line agents for severe multidrug-resistant infections. However, carbapenem-resistant Enterobacterales (CRE) have emerged as a major global health concern4. This resistance is mainly mediated by carbapenemases, including class B metallo-beta-lactamases, contributing to extensively drug-resistant and pan-resistant phenotypes5.
MDR E. coli strains significantly limit therapeutic options and complicate infection management. This underscores the need for alternative treatment strategies such as combination therapies to improve clinical outcomes. Furthermore, these resistant strains are increasingly reported not only from clinical settings but also from food sources such as dairy products. This broadens the epidemiological context of antimicrobial resistance dissemination and supports combination therapy as a potential strategy to mitigate its spread and clinical impact6,7. Aminoglycosides, tigecycline, and polymyxins are commonly used therapeutic agents in MDR infections8. Fosfomycin has also demonstrated activity against ESBL-producing E. coli, particularly in urinary tract infections9.
Fosfomycin, first discovered in 1969, inhibits early bacterial cell wall synthesis and has re-emerged as a promising agent in combination therapy10. Synergistic effects between fosfomycin and aminoglycosides or β-lactams have been reported in MDR organisms11,12. In addition, fosfomycin has been evaluated in large clinical cohorts, demonstrating effectiveness in severe infections when used alone or in combination regimens13. However, most previous studies include heterogeneous bacterial populations, and data specifically addressing carbapenem-resistant E. coli remain limited.
Standard antimicrobial susceptibility testing methods include agar dilution, broth dilution, and disk diffusion. Agar dilution–based combination testing enables quantitative assessment of antimicrobial interactions by determining MICs and calculating FICIs. Alternative methods such as time-kill and gradient diffusion provide dynamic or simplified interaction data but may differ in reproducibility and standardization13,14. This protocol is most suitable for controlled in vitro evaluation of defined antibiotic combinations. However, it does not fully reflect in vivo pharmacokinetic conditions15. Among these, agar dilution is considered a reference method for MIC determination; however, its use in evaluating fosfomycin combination activity against carbapenem-resistant E. coli remains limited.
This study provides a systematic evaluation of fosfomycin-based combinations against clinically confirmed carbapenem-resistant E. coli isolates. A standardized agar dilution checkerboard approach was used, generating direct comparative interaction data under controlled in vitro conditions.
Therefore, the present study tests the hypothesis that fosfomycin in combination with meropenem, gentamicin, or ceftriaxone exhibits in vitro synergistic activity against CRE E. coli isolates.
Access restricted. Please log in or start a trial to view this content.
Review and Perspective
Loading...
$$\rightleftharpoonup{xx}$$
$$\longleftharp{xx}$$,
$$\longrightharp{xx}$$,
Approval for this study was obtained from the Non-Interventional Clinical Research Ethics Committee of Dicle University Faculty of Medicine (Decision No: 100, Date: 01.01.2021).
The protocol below describes a standardized agar dilution checkerboard method for evaluating fosfomycin combination synergy against carbapenem-resistant E. coli isolates.
Collection of isolates
Between 15.02.2021 and 15.12.2021, a total of 60 carbapenem-resistant E. coli strains isolated from various clinical specimens (urine, blood, aspirate, wound, cerebrospinal fluid, and endotracheal aspirate) obtained from inpatients at Dicle University Medical Faculty Hospitals were included. Identification was performed using a MALDI-TOF mass spectrometry system. When multiple isolates were recovered from the same patient, only the first isolate was included. Samples from outpatient clinics and patients hospitalized for less than 48 h were excluded. All carbapenem-resistant E. coli isolates that met the predefined inclusion criteria during the study period were consecutively included; therefore, the sample size was determined by the total number of eligible isolates available, and no a priori sample size calculation was performed. E. coli ATCC 13846 and E. coli ATCC 25922 were used as quality control strains. Quality control strains were tested in parallel with clinical isolates under identical agar dilution conditions. All runs were evaluated by two independent observers. The results of the quality control strains showed expected growth inhibition patterns in all experiments, and no out-of-range or invalid results were observed.
All procedures involving clinical isolates were performed under biosafety level 2 (BSL-2) conditions. Appropriate personal protective equipment (lab coat, gloves, and eye protection) was used, and all manipulations were conducted in a certified biological safety cabinet.
Species identification and antimicrobial susceptibility testing
Carbapenem resistance and antimicrobial susceptibility testing were performed using an automated microbial identification and susceptibility testing system. Carbapenem resistance detected by the system was confirmed by disk diffusion testing with ertapenem, imipenem, and meropenem disks. Zone diameters were interpreted according to EUCAST breakpoints16,17(Table 1). All isolates were preserved in tryptic soy broth containing 20 % glycerol and stored at −20 °C.
Table 1: Zone diameter breakpoints for carbapenems determined for E. coli by disk diffusion test (DDT). Disk diffusion interpretive criteria used for carbapenem susceptibility testing. Please click here to download this Table.
Blood culture bottles suitable for pediatric and aerobic samples were incubated in an automated blood culture system. Positive bottles were subcultured onto EMB agar, 5% Sheep Blood Agar, and Sabouraud Dextrose Agar.
Urine and endotracheal aspirate samples were cultured using quantitative methods, whereas other samples were processed by semi-quantitative methods. Endotracheal aspirate samples were evaluated using Gram staining, and only samples with a Bartlett score > 0 and a Murray score > 3 were considered significant and included18.
For wound samples, Gram staining evaluation was performed according to the Q scoring system; samples with a Q score greater than 0 were included19.
Bacterial isolates were identified by Matrix-Assisted Laser Desorption/Ionization-Time of Flight Mass Spectrometry. A score of ≥2.0 was considered reliable for genus and species-level identification, while scores between 1.7 and 2.0 indicated reliable genus-level identification20,21. Only isolates with scores of ≥2.0 were included.
Determination of fosfomycin susceptibility by the agar dilution method
According to the European Committee on Antimicrobial Susceptibility Testing (EUCAST), the agar dilution method is recommended for determining fosfomycin susceptibility in members of the Enterobacterales family16.
Preparation of fosfomycin and other antibiotic working solutions
For the agar dilution method, powdered fosfomycin disodium salt hydrate was used, and sterile distilled water was used to create the dilution. Eight different dilutions were planned for each antibiotic, ranging from 4-fold below and 3-fold above the MIC limit value determined by EUCAST.
A fosfomycin stock solution (containing 2560 µg/mL fosfomycin) was prepared by dissolving 51.2 mg of sterile fosfomycin disodium, calculated using the volume/weight/potency formula (Volume (mL) = Weight (mg) × Potency (µg/mg)/ Concentration (µg/mL)) and weighed under sterile conditions, in 20 mL of sterile distilled water. The final concentration was planned to be 256 µg/mL, resulting from a 10-fold dilution of the stock solution when MHA was added.
Accordingly, dilution ranges were determined for fosfomycin (2-256 µg/mL), meropenem and gentamicin (0.125–16 µg/mL), and ceftriaxone (0.0625–8 µg/mL).
Dilutions were prepared by serial two-fold dilutions in sterile conical tubes containing 10 mL of sterile distilled water. A total of 7 serial two-fold dilutions were performed in the tubes until a concentration of 2 µg/mL was reached. The remaining 10 mL in the last tube was discarded. This resulted in eight different antibiotic concentrations: 256, 128, 64, 32, 16, 8, 4, and 2 µg/mL.
The same method used for fosfomycin was used for the other antibiotics to prepare the initial stock solution. 1.6 mg of powdered antibiotic was weighed for ceftriaxone and 3.2 mg for meropenem and gentamicin using a precision scale. Seven dilutions were made for each antibiotic. Thus, eight different concentrations of antibiotic dilutions were obtained for meropenem, gentamicin, and ceftriaxone (Figure 1).
Figure 1: Preparation steps of fosfomycin dilution solutions. Serial dilution steps used to obtain working concentrations. Please click here to view a larger version of this figure.
Preparation of agar medium
Cation-adjusted Mueller-Hinton Agar (CAMHA) was prepared following the manufacturer’s instructions. It was sterilized at 121 °C and 1 atm for 15–20 min. using an autoclave, then kept at 45–55 °C in a water bath.
All biological materials were sterilized by autoclaving at 121 °C for 15–20 min before disposal. Antibiotic-containing waste was disposed of in accordance with institutional biosafety and chemical waste protocols.
Critical Step: Maintain agar temperature between 45–55 °C before antibiotic addition. Excessive heat may reduce fosfomycin activity and affect MIC results.
Addition of G6P to the medium
To determine fosfomycin susceptibility, EUCAST recommends adding 25 µg/mL glucose-6-phosphate (G6P) to agar and broth media5. For each liter of agar, 25 mg of G6P was dissolved in 20 mL of MHA, filtered to sterilize, and then mixed into the main agar batch.
Preparation of agar dilution plates
To determine the monotherapy MIC values for fosfomycin, meropenem, gentamicin, and ceftriaxone, one Petri plate was used per dilution (e.g., 8 dilutions for fosfomycin, ranging from 2–256 µg/mL). The antibiotic concentrations were noted on the plates.
For the combination study, the concentrations of fosfomycin and the other antibiotic being investigated were noted on the plates (e.g., fosfomycin 256 µg/mL and meropenem 32 µg/mL). Thus, 64 plates were prepared for each combination, and the concentration of fosfomycin and the other antibiotic used was noted on the plates. A total of 32 meropenem-gentamicinetri plates were prepared for monotherapy and 192 for combination. A standard drawing containing 56 small 1 cm squares was made on a marble surface. All plates were inoculated directly on this drawing. Thus, reading and inoculation were standardized (Figure 2 and Figure 3).
Figure 2: Standard diagram created for inoculation and reading purposes. Standardized scheme used for inoculation and evaluation. Please click here to view a larger version of this figure.
Figure 3: Evaluation method of agar dilution plates. Assessment of bacterial growth for MIC determination. F = fosfomycin; M = meropenem. Please click here to view a larger version of this figure.
Two milliliters of antibiotic solution were mixed with 18 mL of molten CAMHA to obtain the final working concentration. This medium, prepared according to the manufacturer’s package insert, was cooled to 45−55 °C in a hot water bath. One part (2 mL) of the antibiotic mixture was used, and 9 parts (18 mL) of MHA were mixed in the petri dishes (the antibiotic was diluted 1/10). The final agar plate volume was standardized to approximately 20 mL per plate (18 mL agar + 2 mL antibiotic solution), and final antibiotic concentrations were calculated after dilution. In this way, antibiotic concentrations were achieved in the agar dilution plates (2–256 µg/mL for fosfomycin, 0.0625–8 µg/mL for ceftriaxone, and 0.125–16 µg/mL for meropenem and gentamicin). Plates prepared from the medium without antibiotic solution were used as growth controls (Figure 4). The mixtures consisting of medium and antibiotic solution were allowed to solidify at room temperature.
Figure 4: Control medium without antibiotic solution. Medium without antibiotics is used to confirm bacterial viability. Please click here to view a larger version of this figure.
Critical Step: Uneven agar depth or insufficient mixing of antibiotics may lead to inaccurate MIC interpretation and false interaction results.
Preparation and inoculation of isolates
Previously stored pure strains were passaged onto 5% sheep blood agar at room temperature and incubated at 37 °C for 16–24 h. Colonies from fresh cultures were suspended in sterile saline solution, and a 0.5 McFarland turbidity was achieved using an optical reader.
The agar dilution method required a bacterial inoculum of approximately 104 CFU per spot. A 0.5 McFarland value contains approximately 108 CFU/mL of bacteria. A bacterial density of 107 CFU/mL was achieved by diluting the solution 1:10 with sterile saline.
2 µL of each bacterial suspension was collected from the tube using an automatic pipette. This volume corresponds to approximately 2 × 104 CFU per spot. Inoculation was performed on targeted areas of the agar surface, starting with the plate containing the lowest fosfomycin concentration and progressing to the highest.
Thus, inoculation was defined as CFU per spot, with each inoculation point containing approximately 104 CFU. Inoculation of each bacterial suspension was completed within 15 min.
Before inoculating the fosfomycin-containing plates, a growth control plate was inoculated to check for contamination. The inoculation was allowed to dry at room temperature. Incubation was performed at 35–37 °C for 16–24 h under aerobic conditions, and turbidity was measured with a calibrated densitometer.
Excess inoculum density may cause trailing growth and difficulty in endpoint determination. Re-adjust turbidity to 0.5 McFarland and repeat inoculation if confluent growth occurs.
Combination testing
The powdered antibiotics meropenem, fosfomycin, gentamicin, ceftriaxone, and glucose-6-phosphate were used.
Antibiotic amounts were calculated using the volume/weight/potency formula. After preparing the stock solutions, antibiotic solutions were prepared at concentrations of three two-fold dilutions above and four two-fold dilutions below the susceptible MIC range. Dilution ranges were calculated as 2–256 µg/mL for fosfomycin, 0.0625–8 µg/mL for ceftriaxone, and 0.125–16 µg/mL for gentamicin and meropenem (Table 2, Table 3, and Table 4).
Table 2: Preparation of Antibiotic Solutions for Fosfomycin-Gentamicin Combination. Preparation steps and concentrations used for stock and working solutions. Please click here to download this Table.
Table 3: Preparation of Antibiotic Solutions for Fosfomycin-Meropenem Combination. Preparation steps and concentrations used for stock and working solutions. Please click here to download this Table.
Table 4: Preparation of Antibiotic Solutions for Fosfomycin-Ceftriaxone Combination. Preparation steps and concentrations used for stock and working solutions. Please click here to download this Table.
In this study, separate fosfomycin stock solutions were prepared for monotherapy and combination experiments to compensate for differences in final dilution factors during agar-based testing. A standard stock solution was used for monotherapy experiments, whereas an adjusted, higher-concentration stock solution was used for combination experiments to compensate for the additional dilution during antibiotic premixing.
Stock solutions were prepared as the initial step for serial dilution. For fosfomycin, 256 mg of powder was accurately weighed using an analytical balance under sterile conditions and dissolved in 50 mL of sterile distilled water with continuous mixing (magnetic stirrer) until complete dissolution was achieved, resulting in a final concentration of 5120 µg/mL fosfomycin disodium. Separate stock solutions of gentamicin, meropenem, and ceftriaxone were prepared.
For serial two-fold dilutions, seven sterile conical tubes containing 25 mL of sterile distilled water were prepared. From the stock solution, 25 mL was transferred into the first tube using an adjustable pipette, followed by sequential transfer into subsequent tubes to achieve two-fold serial dilutions. Each tube was mixed thoroughly with a vortex mixer for 5–10 s after each transfer step. From the final tube, 25 mL was discarded to maintain equal volumes. In this manner, a dilution series was obtained, yielding a lowest concentration of 40 µg/mL, and a total of eight different fosfomycin concentrations were prepared.
The same procedure was applied for gentamicin, meropenem, and ceftriaxone. However, due to differences in target concentration ranges, 16 mg of gentamicin and meropenem (for a range of 0.125–16 µg/mL) and 8 mg of ceftriaxone (for a range of 0.0625–8 µg/mL) were weighed and dissolved separately in 50 mL sterile distilled water under aseptic conditions at room temperature. Serial two-fold dilutions were performed as described above. Final lowest concentrations were 2.5 µg/mL for gentamicin and meropenem, and 1.25 µg/mL for ceftriaxone. For each antibiotic, eight tubes containing 25 mL of different concentrations were prepared. Since fosfomycin was tested in combination with three antibiotics, a total of 24 fosfomycin dilution tubes (8 × 3) were prepared.
For agar preparation, 18 mL of Mueller–Hinton agar supplemented with glucose-6-phosphate (25 µg/mL) was pre-cooled to 45–50 °C and dispensed into sterile Petri dishes. Subsequently, 1 mL of fosfomycin solution and 1 mL of the second antibiotic solution were added using sterile pipettes. The plates were gently mixed by circular rotation to ensure homogeneous distribution and allowed to solidify at room temperature. Due to equal-volume mixing and agar addition, the antibiotic concentrations were reduced by approximately 9-fold.
For combination testing, 1 mL of the lowest fosfomycin concentration (40 µg/mL) was combined with 1 mL of each of the eight gentamicin dilutions (2.5–320 µg/mL), then added to separate Petri dishes and gently mixed. The same procedure was repeated for the next fosfomycin concentration (80 µg/mL), and continued stepwise up to 5120 µg/mL, ensuring that all possible concentration combinations were tested. A new sterile pipette tip was used for each transfer to prevent cross-contamination. In total, 64 Petri dishes (8 × 8 combinations) were prepared for the fosfomycin–gentamicin combination (Table 5).
The identical procedure was applied to fosfomycin in combination with meropenem and ceftriaxone, resulting in 64 plates for each antibiotic combination (Table 6 and Table 7).
Table 5: Arrangement of Fosfomycin-Gentamicin Combination on MHA Medium. Arrangement of antibiotic concentrations on Mueller–Hinton agar. FF = fosfomycin; GN = gentamicin. Please click here to download this Table.
Table 6: Arrangement of Fosfomycin-Meropenem Combination on MHA Medium. Arrangement of antibiotic concentrations on Mueller–Hinton agar. FF = fosfomycin; MEM = meropenem. Please click here to download this Table.
Table 7: Arrangement of Fosfomycin-Ceftriaxone Combination on MHA Medium. Arrangement of antibiotic concentrations on Mueller–Hinton agar. FF = fosfomycin; CRO = ceftriaxone. Please click here to download this Table.
After preparing the antibiotic mixtures, 18 mL of molten Mueller–Hinton agar (45–50 °C) was added to each plate containing 1 mL of the antibiotic mixture, ensuring that the temperature conditions did not compromise antibiotic activity. This resulted in an approximate 9-fold dilution of the initial antibiotic solution.
For systematic evaluation, plates were arranged in a checkerboard (two-dimensional dilution) format. Fosfomycin concentrations were organized along the vertical axis (rows A–H), with the highest concentration at row A and decreasing concentrations toward row H. The second antibiotic was arranged along the horizontal axis (columns 1–8), with increasing concentrations from column 1 to column 8. Each well-equivalent position, therefore, represented a unique combination of two antibiotic concentrations, allowing comprehensive assessment of interaction effects.
Results evaluation
Minimum Inhibitory Concentration results were evaluated according to EUCAST v.11 recommendations. Growth control plates showed visible bacterial growth, while sterility control plates (uninoculated) showed no growth. The MIC was defined as the lowest concentration at which no visible growth occurred. For combination testing, interaction effects were interpreted using FICI values. Synergy was defined as FICI ≤ 0.5, indifference as > 0.5–4, and antagonism as > 4. For fosfomycin, MIC values ≤ 32 µg/mL were considered susceptible, and those > 32 µg/mL were considered resistant. For gentamicin and meropenem (in meningitis infections), MIC values ≤ 2 µg/mL were considered susceptible, and those > 2 µg/mL were considered resistant. For meropenem (in non-meningitis infections), MIC values ≤ 2 µg/mL were considered susceptible. Values > 8 µg/mL were considered resistant, MIC ≤ 1 µg/mL for ceftriaxone (for non-meningitis infections) were considered susceptible, MIC ≤ 2 µg/mL for ceftriaxone (for meningitis infections) were considered susceptible, and MIC > 1 µg/mL for ceftriaxone (for meningitis infections) were considered resistant16.
Statistical analysis of the study data was performed using statistical software.
Statistics: Categorical variables were expressed as numbers and percentages. Pearson’s chi-square test was used for comparisons in which all expected cell counts were ≥5. Fisher’s exact test was used for analyses with small sample sizes or when any expected cell count was <5. Specifically, the chi-square test was used for overall comparisons of fosfomycin resistance rates, while Fisher’s exact test was used for subgroup analyses of synergistic, additive, and antagonistic interaction rates due to low-frequency distributions in some combinations. A P-value ≤ 0.05 was considered statistically significant.
Interpretation Note: Tiny single colonies within inhibition zones were not considered to be growing unless confluent or reproducible in repeated testing.
Access restricted. Please log in or start a trial to view this content.
Conclusions
Loading...
$$\rightleftharpoonup{xx}$$
$$\longleftharp{xx}$$,
$$\longrightharp{xx}$$,
A total of 60 carbapenem-resistant E. coli isolates were included in the study. All isolates were obtained from patients hospitalized in intensive care units or clinical wards of Dicle University Medical Faculty Hospitals. Adult patients accounted for 83.3% of cases, while pediatric patients accounted for 16.7% (Table 8 and Table 9). Multidrug resistance was observed in 80% (n = 48) of isolates.
Table 8: Distribution of E. coli isolates by hos...
Access restricted. Please log in or start a trial to view this content.
Disclosures
Loading...
$$\rightleftharpoonup{xx}$$
$$\longleftharp{xx}$$,
$$\longrightharp{xx}$$,
The authors declare that they have no conflict of interest.
Acknowledgements
Loading...
$$\rightleftharpoonup{xx}$$
$$\longleftharp{xx}$$,
$$\longrightharp{xx}$$,
This study was supported by Dicle University Scientific Research Projects Coordination Unit (DÜBAP) with project number SBE.21.004.
Access restricted. Please log in or start a trial to view this content.
Reprints and Permissions
Request permission to reuse the text or figures of this JoVE article
Request Permission