Trovafloxacin (Trovan Pfizer), a fluoroquinolone antibiotic, was recently approved by the FDA for human use involving fourteen
indications. At the present time veterinary labeling has not been sought. Its potential advantages over fluoroquinolones currently used in veterinary medicine
include an improved gram-positive spectrum and efficacy against anaerobes7. The margin between minimum arthropathic dose and pediatric therapeutic dose
is greater than that of other fluoroquinolones.2 Therefore, we hypothesize that trovafloxacin will be a safe and efficacious option for oral
administration in the little skate, Raja erinacea. Pharmacokinetics of enrofloxacin, a fluoroquinolone frequently prescribed for the treatment of fish,
have been determined in the red pacu (Colossoma brachypomum).6
Other fluoroquinolones commonly employed in aquatic animal medicine have been reported to cause cartilage damage in weight-bearing joints of
young rapidly growing beagles4,8. Such chondropathology appears common to the quinolones. While the clinical response of animals treated with
fluoroquinolones has been good, we have concerns about our current lack of knowledge on the effects, both expected and adverse, in aquatic animals treated with
newer fluoroquinolones. Concerns about the effects of quinolones on cartilaginous structure and development might well be addressed in a species such as the
little skate in which cartilage is the principal skeletal element. If quinolone-associated lesions, such as erosion and blister formation, develop in articular
cartilage3, this species may prove to be an alternative model for such chondropathies.
Materials and Methods
Thirty locally captured female skates were acclimated to and maintained in an 11,000 L circular pool at a temperature of 13 C, pH of 8.1,
and salinity of 28ppt. On the day prior to initial dosing with trovafloxacin and formal start of the experiment (day -1), specimens were weighed and marked with
an identifying single hole punch at the periphery of the right pectoral fin. Groups of five skates were separated into floating baskets and designated by the
color of the baskets and floats. Blood was drawn via cardiac puncture for baseline CBCs (Unopette method5) and plasma chemistries.
On day 0, skates were dosed by manual administration of either a placebo capsule, ~10 mg/kg (extrapolated therapeutic dosage) as a 15 mg/mL
suspension in a capsule, or ~100 mg/kg (10X therapeutic dosage) as a 50 mg tablet plus additional suspension in a capsule at time 0. Blood samples were taken at
2, 6, 12, and 24 h for pharmacokinetics. Samples were transferred to 1cc lithium heparin microtainer tubes, centrifuged at ~3000 rpm for 10 min, and the plasma
transferred to cryovials and frozen at -70 C until analyzed for trovafloxacin levels using a previously validated HPLC assay.
Skates were subsequently administered a second and third oral dose or placebo at 24 h and 48 h, respectively. At 72 h, blood was drawn again
for pharmacokinetic analysis. In addition, CBC and plasma chemistry analyses were performed on samples from half of the skates of each group. Those individuals
were humanely euthanized at that time with MS-222 @ 600 mg/L. The remaining skates were euthanized at 144 h in order to determine elimination of trovafloxacin
from plasma, as well as CBC and plasma chemistry values at 96 h after the final dose. Liver and both joints on the right side of each animal euthanized were
harvested and submitted for blind histopathologic examination. Dorsoventral, full body radiographs were taken immediately post-mortem.
In addition to determining the pharmacokinetics of trovafloxacin in this representative elasmobranch, radiographic and histopathologic studies
were conducted to better assess whether adult cartilaginous fishes are appropriate models for quinolone induced chondropathies. Based on gross anatomy and
pre-study radiographs, the procondylar and metacondylar pectoral joints appear to be the most mobile, and therefore appropriate, for this study. Liver and joints
were fixed in 10% buffered formalin for 48 hours. The joints were then transferred to Bouin's solution for 48 hours. Frontal section of joints were histologically
examined with H & E stain for surface erosions or deficits, superficial eosinophilia, cleft or vesicle formation, chondrocyte necrosis, and mucopolysaccharide
deposits in ligaments. Representative sections from each therapeutic group were also stained with Von Kossa and toluidine blue. Liver sections were examined for
fatty change, eosinophilic droplet formation, and perivascular mixed infiltrates. Severity of lesions was graded on a scale of 0-5.
Results and Conclusions
Differences in various CBC and plasma chemistry parameters within a group were significant with respect to time. However, differences
between groups were not significant at 0, 72, and 144 h. Mean values for baseline hematologic and plasma chemistry parameters were determined.
Mean plasma concentrations of trovafloxacin for the 10 mg/kg group at 2, 6, 12, 24, 48, 72, and 144 h were 0.015, 0.021, 0.037, 0.045, 0.516,
0.990, and 1.839 mcg/mL, respectively. For the 100 mg/kg group, mean concentrations were 0.020, 0.021, 0.496, 0.927, 3.137, 4.677, and 8.669 mcg/mL, respectively.
The MIC90s for most gram-positive cocci are less than 0.5 mcg/mL, excluding methicillin and ciprofloxacin-resistant S. aureus (1-8 mcg/mL). The
MIC90s for P. aeruginosa, most anaerobes, and most Enterobacteriaceae are 1.0-16.0, 1.0, and 0.5 mcg/mL, respectively.1 C and T
maxmax were not determined because 144 h samples showed a continuing rise in plasma trovafloxacin concentrations, possibly due to the slower metabolism of this
cold water elasmobranch.
Evaluation of post-study radiographs was unremarkable with the exception of a severe metacondylar dislocation with lateral distraction of the
distal bone of one skate in the 100 mg/kg group.
Histopathologic examination of this joint revealed a chronic perichondrial cartilage proliferation and a 0.5 mm focal area of chondrocyte
necrosis, surface fibrillation, and depletion of glycosaminoglycans. No histologic lesions of vesiculation or horizontal cleft formation in the superficial or
middle one-thirds of articular cartilage, consistent morphologically with the quinolone-induced chondrotoxicity documented in mammals, were identified. Foci of
separation of demineralized cartilage in tissue sections (among all groups) were negative for retention of calcium salts using the Von Kossa staining method.
Toluidine blue staining failed to reveal focal depletion of glycosaminoglycans in segments of articular cartilage.
The results of this study indicate that trovafloxacin may be safely and efficaciously administered to a representative elasmobranch at ten
times the prescribed oral dose. Though trovafloxacin did not induce arthropathy in adult skates, we plan future studies with juvenile skates, in which
trovafloxacin might be expected to have a more pronounced effect on developing articular cartilage. Alternatively, fluoroquinolones with lower margins of safety,
such as ofloxacin, may yield more dramatic results. Intravenous injection of C14 radiolabeled trovafloxacin will be utilized for assessment of tissue
distribution and elimination in the future.
This study was supported by a collaborative effort between the Sea Research Foundation in Mystic, Connecticut, and Pfizer Inc. of Groton,
Connecticut. The authors wish to thank the following members of the Mystic Aquarium husbandry staff for their unwavering support in the transport, care, and
handling of the animals: Gregory Charbeneau, Catherine Ellis, Alison Scarratt, Jan Drugge, Philip Fensterer, Kristen Lomme, and Dan Augustino. Amy Tucker of the
Mystic Aquarium laboratory staff spent many eye-strained hours scrutinizing skate CBCs. Gayle Sirpenski and Steve Spencer provided photodocumentation.
Establishing the normal histological parameters of elasmobranch skeletal and articular cartilage would not have been possible without the skills of Dr. Sal Frasca
and Dr. John Burkhardt. Dr. Wayne Carter was instrumental in orchestrating Pfizer's role in the project. Dr. Yuanchao Zhang, Dr. Teresa A Smolarek, Keith A.
Hoffmaster, L. Dean Kendall, and Jeanine M. Block devoted valuable laboratory time to develop the HPLC assay. We also appreciate the efforts of Dr. Donna Zyry at
Pfizer in establishing the literature search. Millstone Nuclear Power Station of Waterford, Connecticut graciously supplied us with skates. Most of all, the
co-mentorship and participation of Dr. Larry Dunn and Dr. David St. Aubin made this internship and project an enjoyable and rewarding educational experience.
1. Brighty KE, TD Gootz. 1997. The chemistry and biological profile of trovafloxacin. J Antimicrobial Chemotherapy 39:Suppl
2. Burkhardt JE. 1996. Review of quinolone arthropathy in the dog. Chemotherapie Journal 5:14-18.
3. Burkhardt JE, MA Hill, WW Carlton, et al. 1990. Histologic and histochemical changes in articular cartilages of immature
beagle dogs dosed with difloxacin, a fluoroquinolone. Vet Pathol 27:162-70.
4. Burkhardt JE, JN Walterspiel, UB Schaad. 1997. Quinolone arthropathy in animals versus children. Clinical Infectious Diseases
5. Jain NC. 1986. Schalm's Veterinary Hematology. Lea & Febiger, Philadelphia, Pp. 261-263.
6. Lewbart G, S Vaden, J Deen, C Manaugh, D Whitt, A Doi, T Smith, K Flammer, 1997. Pharmacokinetics of enrofloxacin in the red pacu
(Colossoma brachypomum) after intramuscular, oral and bath administration. J Vet Pharmacol Therap 20:124-128.
7. Norrby SR. 1997. New fluoroquinolones: towards expanded indications? Current opinion in Infectious Diseases 10:440-443.
8. Yoshida K, K Yabe, X Nishida, N Yamamoto, C Ohshima, M Sekiguchi, K Yamada, K Furuhama, 1998. Pharmacokinetic disposition and
arthropathic potential of oral ofloxacin in dogs. J Vet Pharmaco Therap 21:128-132.