Abstract
Francisella tularensis is a highly infectious gram-negative bacterium that is the etiologic agent of tularemia in animals and humans [1, 2]. Infections with highly virulent F. tularensis strains are lethal in 30%-60% of individuals infected by the inhalation route if not treated with antibiotics, and F. tularensis strains have been weaponized for potential use as a biothreat agent [3-5]. For these reasons, F. tularensis has been designated a Tier 1 Select Agent by the Centers for Disease Control and Prevention (CDC) and the National Institute of Allergy and Infectious Diseases (NIAID), part of the National Institutes of Health. The incidence of pneumonic tularemia worldwide is low; therefore, it is not feasible to conduct clinical efficacy testing of tularemia medical countermeasures (MCMs) in humans. During 2001-2010, 1208 cases of all forms of tularemia were reported in the United States; the median number of cases per year was 126.5 with a range of 90-154 cases per year. These cases typically were spread out geographically and remained sporadic in nature, although small outbreaks have been reported [6]. Average annual incidence was reported as 0.041 cases per 100 000 persons [6]. Even in situations where clinical data have been reported [6-9], it is difficult to describe the disease course due to lack of understanding of timing and route of exposure. In addition, it is useful to note that person-to-person transmission is extremely rare. Thus, there is an inability to conduct human efficacy trials for tularemia treatments, and there is a lack of comprehensive data that describe the course of pneumonic tularemia in humans. In the United States, tularemia MCM efficacy studies can be conducted in accordance with the Animal Rule (21 CFR Parts 314 Subpart I and 21 CFR Part 601 Subpart H, Approval of New Drugs When Human Efficacy Studies Are Not Ethical or Feasible, Food and Drug Administration [FDA]) [10, 11]. Drug or vaccine licensure under the Animal Rule requires availability of at least 1 well characterized animal model displaying pathophysiology similar to disease in humans [10, 11].
The FDA’s Animal Model Qualification Program [12] under the Drug Development Tools Program is a regulatory pathway for qualification of animal models to be used for MCM efficacy testing and approval under the Animal Rule. A qualified animal model is product-independent and may be used for efficacy testing of multiple investigational drugs for the targeted disease or condition. The NIAID and Biomedical Advanced Research and Development Authority worked together to fund supportive studies and generate the regulatory package to establish a qualified animal model for the cynomolgus macaque model of pneumonic tularemia.
Bạn đang xem: Efficacy of Doxycycline and Ciprofloxacin for Treatment of Pneumonic Tularemia in Cynomolgus Macaques
The public document for the qualified animal model for the cynomolgus macaque model of pneumonic tularemia is the Context of Use [13], which describes the characteristics and definitions for use of the model. This includes the critical model parameters, which define the non-human primates (NHPs; cynomolgus macaques) and include the important inclusion and exclusion criteria for their use. It also defines the pathogen strain (Schu S4) characteristics and the means used for aerosol exposure of the animals to the F. tularensis strain. As described [13], Schu S4 is a type A strain that was chosen in consultation with the FDA as the well-characterized challenge material for the qualified model. Finally, it also defines the key end points that define the expected pathophysiology of the disease in the NHPs under these defined conditions. Examples of end points that were established for the pneumonic tularemia model in the cynomolgus macaques included percent survival for placebo/untreated controls, time to fever, time to death, and bacteremia and disease pathology. Other outcomes included changes in hematology and levels of C-reactive protein (CRP). Thus, the qualified animal model defines the experimental setup and expected outcomes to support MCM efficacy testing for tularemia.
In a previous set of studies, efficacy of ciprofloxacin and doxycycline was tested separately using the pneumonic tularemia model in cynomolgus macaques when drug was administered within 6 hours of fever onset using the same dosing schedule as that described here. Treatment at the time of fever onset for both drugs led to 100% survival of drug-treated animals (data not shown). The design and some outcomes of the doxycycline treatment study were briefly described previously [14].
Based on the results of these previous studies, and in consultation with the FDA, the work described here was planned and performed in which drug treatment (either ciprofloxacin or doxycycline) was initiated at both 24 hours and 48 hours post-fever to more closely resemble infections in the field. The goal of this effort was to establish packages of adequate and well-controlled animal efficacy studies using the qualified model for pneumonic tularemia. The studies were intended to provide the primary evidence of effectiveness to support marketing approval of the products.
METHODS
Most of the methods used in these studies were established for the qualified animal model; additional details can be found in the Context of Use document [13] and recent publications describing the model [14-16]. These studies were conducted in compliance with the current version of the US FDA’s Good Laboratory Practice Regulations, 21 CFR Part 58, for the conduct of nonclinical laboratory studies. Protocols were submitted, reviewed, and ultimately approved by the FDA through Special Protocol Assessment.
Test System
Cynomolgus macaques (Vietnamese origin) were procured for the studies from Envigo (now Inotiv), and the animals were screened for a range of inclusion and exclusion criteria as described previously [13, 14]. Animals were aged between 3 and 7 years, weighed between 2.5 and 7 kg, and were equally divided between male and female animals. NHPs were surgically implanted with vascular access ports (VAPs), only for test article administration, and telemetry devices (M00, Data Sciences International), which monitored real-time activity and core body temperature that were used to assess the onsets of fever and hypothermia in real time as previously described [13, 14, 16]. VAPs were locked with anticoagulant and were flushed extensively with saline prior to and after their use for the 1-hour drug infusions and were not used for any of the study blood collections.
Groups of 10 Vietnamese-origin cynomolgus macaques were targeted for each treatment group across all studies (see Table 1 for treatment groups). Ciprofloxacin and doxycycline were tested independently in 2 separate studies, and each efficacy study was performed over 2 separate challenge days (each with 15 animals) such that treatment groups were equally represented on both challenge days. Because of issues with patency of VAPs, 1 animal was excluded from the ciprofloxacin efficacy study such that only 9 animals were in the group treated with ciprofloxacin at 24 hours after onset of fever. Otherwise, animals were released for study and aerosol challenge in the BioSafety Level-3 (BSL-3) containment suite after at least a 30-day quarantine and monitoring of health status prestudy as described [13, 14]. Baseline telemetry data were collected for 14 days prior to challenge during prestudy monitoring and acclimation to the BSL-3. Veterinary observations and exams, including clinical chemistry and hematology values collected both 2 weeks and 1 week prior to challenge, were performed to confirm animal health status.
On each day of the F. tularensis challenge (2 challenge days per drug efficacy study), 15 NHPs were exposed to aerosolized Schu S4 (NR- 10492; from BEI Resources) as described previously [13, 14, 16] at a target dose of 1000 colony-forming units (CFU; approximately 50× Lethal Dose 50% (LD50)) using fresh vials for each exposure without freeze-thaw. Animals were randomized to days of challenge by body weight and then to treatment groups upon fever onset after aerosol exposure.
Drug Administration
The pharmacokinetics (PKs) of ciprofloxacin were modeled based on studies in healthy macaques, and dosing was designed to meet the requirements of the Animal Rule targeting blood levels in humans treated with 400 mg intravenous (IV), twice daily (BID) (as decided based on discussion with FDA and CDC). Ciprofloxacin for injection (18 mg/kg at 9.0 mL/kg) or 5% dextrose was administered by slow IV infusion for 60 minutes (±10 minutes) twice daily for 14 consecutive days starting 24 or 48 hours ±2 hours post-confirmed fever (final elevated temperature reading).
The PKs of doxycycline were modeled based on preliminary PK studies in healthy macaques, and dosing was designed to meet the requirements of the Animal Rule targeting known data regarding blood levels in humans treated with 100 mg IV, BID (as decided based on discussion with FDA and CDC). Thus, for the first doxycycline dose, 8 mg/kg (approximately 15.0 mL/kg) doxycycline for injection was administered for the first infusion (ie, dose 1), followed by 10 mg/kg (approximately15.0 mL/kg) doxycycline for injection for the next 2 infusions (ie, doses 2 and 3) and 15 mg/kg (approximately 15.0 mL/kg) doxycycline for injection twice daily for the remaining 25 infusions. Doxycycline for injection or 0.9% sodium chloride was administered by slow IV infusion for 60 minutes (±10 minutes) twice daily for 14 consecutive days starting 24 or 48 hours ±2 hours post-confirmed fever (final elevated temperature reading). In order to maintain the blind, animals receiving the 0.9% sodium chloride for injection vehicle control received a volume per kilogram equivalent to the volume per kilogram of doxycycline being infused, and the volume and rate calculated for each animal remained constant for all 30 infusions.
Xem thêm : Cuándo hacer el test de embarazo y acertar
Efforts were made to ensure that staff who came in contact with animals were unaware if animals received the test article or vehicle, including the veterinarian and study director. An independent scientist not involved with performance of the study maintained the blinding information, and no treatment information was provided to the staff or study director until the study was unblinded. The study was unblinded (except for the pathologist) when the independent scientist provided the randomization information to the study director at the end of the in-life period and after the completion of the bacteriological blood and tissue culture (culture plates read) end of in-life samples. The study pathologist remained blinded until the reading of the slides for each animal was complete and documented.
In both studies, all animals received their first infusion 24 hours ±2 hours post-confirmed fever, and subsequent infusions were performed 12 hours ±3 hours based on the start time of the previous infusion. The initiation of drug treatment at 24 or 48 hours was based on the treatment group the animal was randomized to at the time of trigger (fever onset). In addition, all animals, regardless of randomization to the 24-hour or 48-hour treatment initiation, received 2 additional vehicle infusions in order to maintain the blind. Animals randomized to a treatment initiated at 48 hours post-confirmed fever received vehicle at 24 hours and 36 hours post-confirmed fever. Similarly, animals randomized to a treatment initiated at 24 hours post-confirmed fever received vehicle for 2 additional doses past their treatment schedule so that they received an infusion consistent with the 48-hour post-confirmed fever group. Therefore, all animals received 30 infusions.
Blood Collection and Disease Monitoring
Blood draws for assessment of bioanalytical, quantitative polymerase chain reaction (qPCR), hematology, clinical chemistry, bacteremia, and serology (by tube agglutination assay) were collected from the most readily accessible vessel and were not collected from the VAP used to infuse the treatment. Blood was processed to serum (for serology and clinical chemistry) or plasma (for hematology and qPCR), and assays were performed as previously described [13, 14]. Bacteremia was measured by plate culture with blood samples collected into sodium polyanethol sulfonate collection tubes as described [13, 14]. Antibiotic levels were measured from plasma samples using validated High Performance Liquid Chromatography (HPLC) assays.
Animals were monitored, and clinical observations were recorded 3 times per day after challenge until the time at which animals were no longer febrile. Note that in-life of the doxycycline efficacy study was extended to 47 days to monitor for possible signs of disease relapse after cessation of bacteriostatic antibiotic treatment. During observations, veterinary staff determined that macaques should be euthanized if the animals displayed significant respiratory abnormalities, other significant symptoms (eg, seizures, paralysis, prostration, or unresponsiveness, ie, the macaque laid down, did not get up when approached, and was unresponsive to external stimuli), or hypothermia, which was defined as 9 consecutive (approximately 2 hours) body temperature readings that were ≥1.5°C less than the lowest of all baseline 30-second sampling time averages (in nonanesthetized animals).
Animals that were euthanized or succumbed to disease underwent gross necropsy and tissue collection to be used for histopathology and assessment of tissue bacterial burden as described previously [13, 14].
RESULTS
Aerosol Exposure and Survival
The average aerosol exposure dose of F. tularensis Schu S4 for all animals challenged was 2535 CFU in the ciprofloxacin efficacy study and 1644 CFU in the doxycycline study. Following aerosol exposure to Schu S4, the course of disease developed as anticipated, given the experiences with the qualified model [13, 14, 16]. For example, the time to onset of fever was similar across treatment groups and between the 2 studies (Table 1). This was accompanied by changes in the secondary end points CRP and hematology measures (white blood cell, neutrophil, and lymphocyte counts) prior to initiation of treatment (data not shown) that were consistent with those that had been previously reported in the model [13, 14].
Initiation of antibiotic treatment at the predetermined times after fever rapidly reversed pathophysiologic signs of tularemia, including elevated body temperature returning to diurnal rhythm within 2-3 days (not shown). However, in placebo-treated animals, 8 of 10 NHPs were euthanized or succumbed to disease in both studies (Figure 1). Time to death in those placebo-treated animals that succumbed to disease was consistent with previous studies [13, 14, 16]; mean time to death in the ciprofloxacin study was 11.8 ±5.0 days (median = 11.2) and 12.5 ±4.9 days (median = 9.0) in the doxycycline study. In contrast, antibiotic treatment with either ciprofloxacin or doxycycline at both time points led to almost 100% survival of treated animals (Figure 1) with only 1 antibiotic-treated animal euthanized (at study day 17), which came from the group treated with doxycycline at 24 hours after fever onset.
Detection of F. tularensis in Animals Exposed to Schu S4
Although bacteremia is not typically a robust indicator of infection for tularemia [14], 40% (3 of 8) of placebo-treated animals in the doxycycline study were culture-positive at terminal euthanasia, while 60% (5 of 8) were positive in the placebo group in the ciprofloxacin study. The kinetics of the proportion of animals that were positive for bacteremia in the 2 studies are shown in Figures 2A and 2C. Even though a low percentage of positive animals were bacteremic, treatment with either doxycycline (Figure 2A) or ciprofloxacin (Figure 2C) decreased the number of positive animals, and bacteremia was only detected just prior to or early after initiation of treatment. In the doxycycline study, there were no positive animals after treatment was initiated, including the 1 animal that succumbed prior to study termination. Similarly, in the ciprofloxacin study, no animals were found to be bacteremic after study day 6, which, depending on treatment group and exact onset of fever, would have been 1-2 days after initiation of antibiotic infusions.
Tissue Bacterial Burden
In contrast, F. tularensis is readily detected from tissues of infected NHPs using plate culture (Table 2, Table 3). Upon euthanasia and necropsy, a range of tissues were harvested. In placebo-treated animals, most, if not all, of the animals that succumbed to disease had detectable bacteria in tissues such as the lung, spleen, liver, and some of the draining lymph nodes. Across the 2 studies, about 50% of placebo-treated animals had measurable bacterial burden in brain as well. In the 2 studies, there were even culturable bacteria in tissues albeit below the lower level of quantitation (LLOQ) from placebo survivor animals; lung and mesenteric lymph node (MLN) in the ciprofloxacin study and both spleen and liver in the doxycycline study.
Antibiotic treatment, as might be expected from the survival of treated animals, markedly inhibited bacterial spread and productive infection of peripheral tissues. In the animals treated with ciprofloxacin, irrespective of treatment group, there were no tissues where measurable bacteria were found (Table 2).
Doxycycline treatment, however, did not have such a sterilizing effect in treated animals (Table 3). In the group in which treatment was initiated 24 hours after fever, none of the surviving animals had detectable tissue bacterial burden. The 1 doxycycline-treated animal from this group that was euthanized prior to the end of the study did show some small evidence of tissue bacterial burden, but only in the MLN and retropharyngeal lymph node and at levels below the LLOQ. On the other hand, even though 100% of the animals survived, there were 3 animals with positive tissues in the group of animals treated with doxycycline 48 hours after onset of fever. One was positive in both spleen and bronchial lymph node (BLN), including levels above the LLOQ in a gross lesion collected at scheduled necropsy on study day 47. Another was positive in spleen, liver, and BLN at levels below the LLOQ, while the third animal was positive in both MLNs and BLNs, with levels in the BLN greater than the LLOQ. Thus, surviving animals in both the placebo and doxycycline-treated groups had measurable bacteria in their tissues upon euthanasia at the end of the study.
qPCR Detection of Bacterial DNA
Xem thêm : Your Guide To The Best and Worst Foods For Gout
Although it does not indicate the presence of live bacteria in blood, the validated qPCR assay was used in these studies to track infection and was found to be a more sensitive indicator than bacteremia by culture (Figures 2B and 2D). Early in the disease, 4 of the 6 treatment groups demonstrated 100% of animals with positive qPCR signals (above the lower level of detection), while the other 2 were 80% (doxycycline, 24-hour delayed treatment) or 90% (ciprofloxacin, 24-hour delayed treatment). Similar to bacteremia, treatment with antibiotics decreased the percentage of positive animals, although positive samples were still observed past study day 12 for both antibiotics. However, by study day 28, all antibiotic-treated animals were negative for qPCR. While the placebo survivors from the 2 studies remained qPCR-positive for a more sustained period of time, all placebo survivors as well as the doxycycline-treated animal that succumbed were PCR-negative at the time of euthanasia.
DISCUSSION
Following the guidance set out in the Context of Use for the qualified animal model for pneumonic tularemia in the cynomolgus macaque [13], these studies were designed to test the efficacy of the antibiotics doxycycline and ciprofloxacin in separate studies. In order to model the clinical setting more closely, antibiotic treatment was initiated either 24 or 48 hours after onset of fever, which was used as a clinical sign of disease onset. The results indicate that both ciprofloxacin and doxycycline were highly efficacious, leading to nearly 100% survival under all treatment scenarios. There were some differences between the bactericidal drug ciprofloxacin and the bacteriostatic doxycycline. However, using treatment regimens that model approved human dosing, there is confidence that these drugs would be useful in treatment of pneumonic tularemia.
Based on the Context of Use, there are certain expected outcomes and end points for well-controlled studies; to a large extent, these studies met those expectations. It is unclear why 2 of the 10 placebo-treated animals survived in each of the 2 separate studies, although the administration of the volumes of saline or dextrose in the placebo animals could be considered supportive measures that may have promoted survival compared with that described for animals in natural history studies [13]. The inhaled dose of Schu S4 and time to fever were not different in the surviving animals compared with other animals in the 2 studies, suggesting that exposure and early stages of disease onset were not markedly different from what was experienced by those animals that succumbed to disease.
Antibiotic dosages were established based on the principles of the Animal Rule in which the exposures in macaques (area under curve and/or Cmax) should be lower than that established in humans. Based on modeling and control PK studies in uninfected animals, dosing was designed (in consultation with both the FDA and CDC) to approach that for the approved dosing in humans for doxycycline (100 mg IV, BID) and ciprofloxacin (400 mg IV, BID). In the case of doxycycline, because of different half-lives in humans (approximately 20 hours) and NHPs (≍ 3.5 hours), the PK modeling (again in discussion with FDA and CDC) indicated the need to escalate the administered dose in NHPs to account for drug accumulation in humans prior to steady state. Furthermore, while loading doses are often used for doxycycline [17], matching the approved dosing according to the Animal Rule dictated targeting the NHP dosing to the approved 100-mg IV, BID treatment regimen in humans. At that dosing, the results suggest that both doxycycline and ciprofloxacin are highly efficacious in the treatment of pneumonic tularemia when administered either at 24 or 48 hours after disease onset. It should be noted that these studies were performed using IV dosing with ciprofloxacin and doxycycline based on discussions with the FDA and the desire to address treatment of severe disease. Nevertheless, while both of these drugs are well absorbed when administered orally in humans, it is possible that during severe tularemia infections, oral therapy might not be as efficacious.
With respect to doxycycline, 1 animal in the group treated at 24 hours post-fever was euthanized prior to the end of study. However, it is unclear if the animal died from complications associated with tularemia, and the cause of death of the animal remains undefined. Aerosol exposure, fever onset, seroconversion, and clinical observations were consistent with other animals in the studies and historical data, so there is confidence that aerosol exposure was done properly, although this animal had unusually frequent emesis. The markers of bacteria in this macaque were more like those in antibiotic-treated animals than in the placebo controls. Bacteremia (culture-negative) and PCR detection of bacterial DNA (only 2 time points at below the LLOQ) were low or absent at all time points measured, and there were only 2 lymph node (LN) culture positive upon necropsy, but no other tissues were positive. Responses to doxycycline treatment were as expected with resolution of fever and return of body temperature to normal diurnal rhythm within 2-3 days after treatment; however, the animal succumbed prior to cessation of treatment (study day 17). Pathologic examination recorded that… “disease pathology was not consistent with tularemia as a cause of death…and…a cause of death could not be established, based on histopathology.” Thus, even the animal that was euthanized from the antibiotic-treated group showed evidence of antibiotic efficacy, but the cause of death was unclear.
After aerosol exposure, there was clear dissemination of the bacteria to peripheral tissues in all animals, including spleen, liver, and kidney, and penetration to and productive infection of the brain were observed in approximately half of the animals in both studies. An interesting observation in the doxycycline-treated animals was that, in spite of the 100% survival in the group treated at 48 hours post-fever, some surviving animals still exhibited the presence of live bacteria in peripheral tissues when euthanized on study day 47. This possibility, plus the evidence that disease relapse occurs with doxycycline-treated individuals who are given shorter or suboptimal courses of treatment [18, 19], was part of the rationale for extending the in-life for the doxycycline study out to 47 days, targeting 28 days after cessation of antibiotic treatment. Nevertheless, none of the animals displayed other clinical signs of disease (clinical observations) or fever, suggesting the animals had controlled the infection.
This was in contrast to macaques treated with the bactericidal antibiotic ciprofloxacin, where no animals had evidence of residual tissue bacterial burden. Interestingly, there were placebo survivor animals in both studies that also had low-level tissue burden, suggesting that in the presence or absence of antibiotic treatment, there are inherent protective mechanisms by which animals can survive and not exhibit signs of disease with viable bacteria. All of these animals had strong antibody titers (as measured by tissue agglutination [14]) within 10 days after challenge and out to day 47 when euthanized, but the levels either before or after challenge were not different from levels in animals that succumbed to disease (data not shown).
In conclusion, these well-controlled and well-documented antibiotic efficacy studies, following the guidelines of the Qualified Animal Model and the Animal Rule, demonstrate the efficacy of the antibiotics doxycycline and ciprofloxacin when administered at levels consistent with humanized dosing to cynomolgus macaques exposed to aerosolized F. tularensis Schu S4. Note that across multiple strains and biovars of F. tularensis, there are varied sensitivities to both ciprofloxacin and doxycycline compared with Schu S4 (by minimum inhibitory concentration). Thus, it is possible that some strains might not respond to treatment in humans as favorably as has been demonstrated in this study of Schu S4 in macaques. Nevertheless, because of the long history of these drugs and the established toxicology literature for them when used in humans, the results demonstrate that these drugs should be efficacious in humans when using approved dosing schedules.
Notes
Acknowledgments. The authors thank the staff at the Battelle Memorial Institute for their hard work and expertise in performing the Good Laboratory Practice studies, particularly the study director for both studies, Lisa Henning, and the Quality Assurance Staff, including Nancy Tirano, Sherrie Wade, and Jana Ruth. They also thank all of the contractors and National Institute of Allergy and Infectious Diseases (NIAID) and US Food and Drug Administration staff who contributed to the Animal Model Qualification package and approval.
Financial support. This work was supported by contract HHSN272201700012I_HHSN27200007 awarded to the Battelle Memorial Institute from the NIAID.
Supplement sponsorship. This article appears as part of the supplement “Tularemia: Update on Treatment and Clinical Findings,” sponsored by the Centers for Disease Control and Prevention.
References
Author notes
Nguồn: https://vuihoctienghan.edu.vn
Danh mục: Info
