Introduction

A novel technique for sex identification of immature chelonians has been recently described.¹ In brief, this technique consists of the visualization of gonads by cystoscopy, through the thin, transparent urinary bladder wall. As opposed to the previously described coelioscopic technique, the cystoscopic one does not require surgical access of the coelom.² Considering the relatively large size of the chelonian bladder, cystoscopy (beyond identification of gonads) can allow visualization of most coelomic viscera.^1,3 In this report, we show for the first time the clinical usefulness of this technique for the diagnostic evaluation of the coelom.

Case Report

A 55 g, 7 mo old Chinese three-striped box turtle (Cuora trifasciata) was presented to the Clinica per Animali Esotici (Rome, Italy) with a 2 wk history of lethargy and anorexia. Upon physical examination, the animal appeared in poor body condition, the eyes were closed, and it was unresponsive to external stimuli. The owner reported that the last feces produced by the turtle were yellowish. In the same breeding facility, five other freshwater turtles had died within the last 2 mo presenting with similar clinical signs (i.e., anorexia, lethargy and yellow feces). The turtles were two Philippine pond turtles (Siebenrockiella leytensis), two Chinese yellow-margined box turtles (Cuora flavomarginata), and one keeled box turtle (Pyxidea mouhotii) housed in different terraquaria in the same room. Each container was equipped with a filter and a mercury vapor lamp^a operating 12 hr per day. Substrate consisted of bark and sphagnum. The turtles were fed invertebrates, leafy vegetables (mainly Chicorium sp.), and, rarely, a commercial pellet diet^b. All six turtles had been treated by the owner with intramuscular 5 mg/kg enrofloxacin^c every 24 hr and 2 mg/kg marbofloxacin^d every 48 hr for 2 wk, without clinical improvement. One of the previously deceased turtles, the Pyxidea mouhotii, was necropsied immediately after its death using standard techniques.^4,5 On gross examination, the animal was emaciated and had several variably sized and roughly round, friable, white-to-yellow areas of discoloration within the hepatic parenchyma (Figure 1A). Samples of the liver, intestine, stomach, pancreas, kidneys, and lungs were fixed in 10% buffered formalin and routinely processed for histology. For microbiology testing, samples were collected from sections of the liver including the friable, whitish areas by use of the sear and stab technique and shipped in bacteriology transport media to the diagnostic laboratory for analysis.⁴ The sample was inoculated on blood agar and MacConkey agar plates at 30° and 37°C in aerobic condition for at least 48 hr. Isolated bacteria were identified with standard biochemical methods. The standard Kirby-Bauer disc diffusion test was performed, using Mueller-Hinton agar and incubation at 30° and 37°C for 48 hr. Samples of the liver were impressed on glass slides stained with Ziehl-Neelsen stain in order to investigate the possibility of mycobacteria or other Ziehl-Neelsen positive microorganisms. The remainder of the liver was frozen at −80°C for further analyses.

View Figure

• Download as Powerpoint
• Download Figure

Histological examination revealed the presence of granulomas and extensive necrosis of the liver (Figure 1B). These lesions often extended to the coelomic surface of the liver and adjacent organs. Within one of the histologic sections of the pancreatic parenchyma, there was a relatively large aggregate of inflammatory cells composed of macrophages, heterophils, and lesser numbers of lymphocytes admixed with mild to moderate amounts of necrosis and necrotic cellular debris. Extravasated erythrocytes were also present. No significant pathological changes were detected within other organs. No Ziehl-Neelsen positive bacteria were identified in any sections.

Culture of the hepatic lesions showed the presence of an antibiotic-resistant Escherichia coli. Serotyping of the involved strain by rapid agglutination method was attempted. Because the gross pathologic features of the lesions resembled coligranulomatosis in avians, the panel of antisera employed for identification of common E. coli in farm species was used. The sample was negative for several single antisera. Gross and microscopic pathology, in addition to the culture of the hepatic lesions, suggested that E. coli was the most likely etiological agent of this pathologic process, even though a concurrent/predisposing viral etiology could not be completely ruled out.

Total body radiographs of the Cuora trifasciata in dorso-ventral, antero-posterior, and lateral view showed adequate inflation of the lungs and the presence of a small amount of gas within the gastric lumen. In a small amount of feces (1–2 g) obtained through cloacal lavage, no parasites were found. A blood sample (0.06 mL) was collected with a 25-gauge needle into 1-mL plastic syringe from the dorsal tail vein. A blood smear was prepared immediately using blood without anticoagulant, and the remainder of the blood was immediately placed in a 32 × 0.8 mm heparinized capillary tube^e for determination of hematocrit. The blood smear was stained with modified Wright's stain and differential leukocyte count was performed as described elsewhere.⁶ The hematocrit was inside the reference interval described in a species of the same genus (20%, reference interval: 15.5–28%).⁶ Complete blood count showed mild leukocytosis (9.5 × 10⁹/L, reference interval: 0.76–8.89 × 10⁹/L), with heterophilia (62%, reference interval: 26–53%), normal eosinophils (10%, reference interval: 0–26.5%) normal monocytes (4%, reference interval: 3.5–9.5%), decreased lymphocytes (13% reference interval: 14.5–32.5%) decreased basophils (11%, reference interval: 14.5–29.5%).⁶ Atypical blood cells were not found.

Considering the necropsy results of the other turtle, coelomic cystoscopic evaluation was performed. The turtle was premedicated with oral 10 mg/kg tramadol^f administered 6 hr before the procedure and sedated with 0.1 mg/kg intramuscular injection of dexmedetomidine^g into the pectoralis major muscle of the left pectoral limb 20 min prior to the procedure.⁷ During the procedure, the operator held the animal in ventral recumbency with the left hand, while the right hand maneuvered a 9.5 French, 30° viewing rigid diagnostic endoscope^h. The endoscope was inserted through the cloaca and directed toward the urethral opening, which is located ventral to the rectum.^8,9 Access to the urethral sphincter was obtained by infusing warm (30°C) lactated Ringer's solution (1 drop every 3 to 4 sec) to allow distension of the urethral opening. Once the access to the urinary bladder was gained, the endoscope was directed cranially to permit visualization of the stomach, pancreas, liver, and the heart.

The liver appeared dark red to brown with multifocal irregularly shaped white-tan and roughly round areas of discoloration on the capsular surface (Figure 2A). The liver margins were scalloped because of these lesions. During the cystoscopy, the ovaries were also visualized by directing the endoscope dorso-laterally. After completion of the cystoscopic procedure, sedation was reversed by intramuscular injection of 0.5 mg/kg atipamezoleⁱ into the pectoralis major muscle of the right pectoral limb, and the turtle was placed in a small warm enclosure within the reptile ward where she recovered uneventfully from the cystoscopic procedure.

View Figure

• Download as Powerpoint
• Download Figure

Based on the macroscopic appearance of the liver, potential differential diagnoses were: granulomatous/necrotizing hepatitis, multifocal hepatic mineralization, hepatic fibrosis, and primary or metastatic neoplastic lesions. Neoplastic lesions were considered unlikely due to the age of the turtle. Considering that the gross lesions were similar to the ones detected in the Pyxidea mouhotii previously necropsied and because of the presence of heterophilia, an E. coli infection was suspected to be the most likely pathological condition. Antibiotic treatment based on sensitivity testing performed on samples from the Pyxidea mouhotii was initiated (20 mg/kg intramuscular injection of ceftazidime^j every 72 hr). Treatment resulted in a rapid resolution of the clinical signs, with gain of appetite 4 days after the first administration. Antibiotic treatment was discontinued 8 days after resolution of symptoms, i.e., overall 12 days of antibiotic treatment were used.

The Cuora trifasciata was re-examined 4 mo after the treatment and appeared in good body condition with increased body weight. A recheck cystoscopy demonstrated the macroscopic absence of hepatic lesions (Figure 2B). The liver margins were smooth and the liver color was normal.

Discussion

Medical approach to young chelonians is often frustrating. The small size of posthatchling chelonians and their anatomical peculiarities (i.e., presence of carapace) complicate identification of the diseased organ systems based upon physical examination alone. Furthermore, size and anatomical characteristics complicate the applications of diagnostic imaging (i.e., radiology and ultrasonography) and limit the laboratory diagnostics (i.e., hematology and serum chemistry). In light of the latter statement, endoscopic techniques allowing macroscopic examination of the viscera would provide a large amount of diagnostic information. Coelioscopy has been widely used for diagnostic examination of the viscera of chelonians.^2,10,11 More recently, cystoscopy was used to examine coelomic viscera: the reproductive tract, liver, intestine, lungs, coelomic wall, left and right kidneys, left and right adrenal glands, stomach, heart, caudal vena cava, and pancreas were visible in small chelonians (i.e., 23.7 to 57.8 g) through the thin urinary bladder wall.¹

In the present case, we describe the use of cystoscopy to visualise a diseased liver. An advantage of the cystoscopic approach compared to the coelioscopic one is that there is no invasion of the coelomic cavity, decreasing chances of post operative coelomitis or haemorrhage related to coelomic access. Potentially, cystoscopy may also be a faster procedure. In posthatchling Hermann's tortoises, 90 sec were necessary on average to visualize gonads through cystoscopy.¹ Furthermore, in the present case, cystoscopy was employed in order to avoid surgical anesthesia. Given the limitations for basic anesthetic monitoring due to the size of the turtle (55 g), as well as its poor health status, we considered general anesthesia extremely dangerous. Although coelioscopy has been sporadically performed in conscious chelonians, recent evidence suggests that anesthesia and analgesia are necessary.^2,10,11 In fact, in a randomized trials, turtles undergoing coelioscopy under general anesthesia presented a significantly better anesthetic score than turtles that received local anesthesia alone.² In the current case, we provided long-lasting analgesia with a μ-opioid receptor agonist and mild sedation with a α ₂-adrenergic agonist.^7,12 It should be considered that the systemic effects of dexmedetomidine alone are not described and are potentially dangerous. Its R-enantiomer, medetomidine, decreased the heart rate, respiratory rate, blood pressure, and PO₂ in desert tortoises (Gopherus agassizii).¹² Similarly, medetomidine-ketamine combinations administered intravenously in desert tortoises provoked moderate hypoventilation.¹³ Reversal of an α ₂-adrenergic agonist in tortoises should be made cautiously; although atipamezole hastens recovery time, its intravenous administration may induce severe hypotension.^12,13

The main limitation of cystoscopy is that it does not permit biopsy or sample collection of the visualized structures. Therefore, histopathological, cytological, and microbiological analyses cannot be performed. It should thus be considered that the implication of E. coli as the causal organism of the liver disease in the Cuora trifasciata is a supposition. Although the lack of diagnostic information other than the visual appearance of the organs limits the usefulness of this technique, rapidity and minimal invasiveness make this technique useful to follow up clinical lesions in affected animals or to screen the coelomic cavity in cases in which more invasive technique are not feasible.

An additional concern related to the use of cystoscopy is the risk of iatrogenic damage. The urethra of chelonians is short, wide, and largely distensible; in fact, it is not uncommon for eggs to reach the urinary bladder in a retrograde manner through the urethral opening.⁸ Therefore, the access of a small diagnostic endoscope (9.5 French, approximately 3.2 mm, of external diameter in the present case) through liquid instillation is easy and minimally invasive.¹ In addition, no evidence of bladder damage was found after cystoscopy in chelonians.¹ Although the risks of cloaca, urethra, and bladder trauma, as well as further iatrogenic abdominal damage, are not insignificant during cystoscopy, they are probably rare-to-uncommon events if appropriate technique is employed.

E. coli is a gram-negative bacterium, belonging to the Enterobacteriaceae, which acts as a primary or secondary cause of disease in a multitude of species, including birds and reptiles.^14,15,16 In birds, different manifestations of E. coli infection have been described including: septicaemia, enteritis, granulomas, omphalitis, sinusitis, airsacculitis, arthritis/synovitis, peritonitis, pericarditis, cellulitis, and swollen head syndrome.¹⁴ E. coli is frequently found in the intestinal microbiota of reptiles, being isolated in 50% of animals studied.¹⁷ Although the pathogenesis of E. coli is poorly understood, there is some evidence that stress and a resulting bacteremia are essential factors for the development of clinical disease and mortality in birds.¹⁴ The current manuscript describes the occurrence of a pathogen in a turtle breeding facility, that—although different—shares several characteristics with intensive livestock farming (e.g., limited space for each animal, frequent animal movements, a number of animals in a single room). It is possible that such conditions may have contributed to the development of the disease. Interestingly, a similar condition has been described in free-ranging marine turtles. In one survey, approximately 14% of stranded sea turtles were affected by granulomatous hepatitis. Together with necrotizing hepatitis, they were the most commonly observed hepatic lesion.¹⁵ Furthermore, E. coli was isolated on 31% (5/16) of the cases of necrotizing-granulomatous hepatitis.¹⁵ Granulomatous hepatitis affecting sea turtles was similar to that observed in the semiterrestrial Pyxidea mouhotii; it was grossly characterized by small, yellowish-whitish granulomas scattered randomly throughout the liver. Microscopically, those granulomas were characterized by necrotic centers surrounded by macrophages and multinucleated giant cells.¹⁵ The authors attributed those hepatic lesions to multisystemic septicemic status.¹⁵

Conclusion

In conclusion, we describe for the first time the use of cystoscopy to diagnose a liver lesion in a chelonian without the use of surgical access of the coelom. In the present case, a definitive, etiological diagnosis was not achieved. Nevertheless, a clear identification of the diseased system through cystoscopy was possible allowing a more informed therapy. The main limitation of this technique compared with the coelioscopic approach remains the impossibility to obtain diagnostic biopsies. Considering how challenging the medical approach to posthatchling chelonians is, a technique allowing rapid and non-invasive localization of diseased organ systems is extremely valuable to the clinician.