Clinical and Pathological Findings in a Red-Tailed Boa Constrictor (Boa constrictor constrictor) with a Primary Neural Neoplasm within the Diencephalon and Mesencephalon

A 22-yr-old captive-bred male red-tailed boa constrictor (Boa constrictor constrictor) was evaluated for abnormal head position for 1-day duration. Husbandry of the snake was considered adequate and included a 4 × 28–foot enclosure with ceramic heat sources to maintain a thermal gradient of 33.3°C (92°F) to 22.2°C (72°F). The diet of the snake included frozen-thawed adult rats (Rattus norvegicus) every 3–4 wk. The snake was a part of a large collection of reptiles including other snakes, chelonians, and lizards. No new snakes had been added to the collection in the previous 16 yr. The cage mate (same species) of the snake had died 1 yr prior; a full postmortem evaluation in that animal diagnosed lymphoma without any histopathologic findings supportive of inclusion body disease of boids.

On initial physical examination, the snake exhibited a 90° head turn to the right, and when stimulated during handling or feeding, the snake would roll to the right in tight curls. No other physical examination abnormalities were noted at that time. A complete blood count and plasma biochemistry profile were unremarkable based upon reference values for the species (Chiodini and Sundberg 1982; Diethelm and Stein, 2006). A gamma globulin level of 0.47 g/dl obtained by serum protein electrophoresis was considered indicative of hypogammaglobulinemia based on a published reference interval of 1.3–2.2 g/dl in a related subspecies (Boa constrictor amarali) (Silva et al., 2011). Testing for inclusion body disease of boids, including complete blood count scanning for cytologic evidence and PCR testing of the blood and mucosal swabs for snake arenavirus, were negative. The snake was treated with a course of enrofloxacin (10 mg/kg, IM, q 48 h, diluted in fluids, Baytril©, Pfizer, New York, NY) and meloxicam (0.1 mg/kg, IM q 24 h, Metacam, Boehringer Ingelheim Vetmedica, Inc., St. Joseph, MO) for 30 days with no improvement in clinical signs. During this time, the snake was supported with intermittent subcutaneous fluid therapy and warm water soaking to maintain hydration.

Approximately 2.5 months after initial presentation, the snake appeared to be showing mild clinical improvement. The head turn was less prominent and the tight rolling occurred with less frequency. One month later (3.5 months after initial presentation), the snake began accepting food on its own. However, clinical signs of the same initial severity returned 2 months later (5.5 months after initial presentation) and the same previously described supportive care was applied for the next 4 months, during which time the snake improved slowly but never regained the ability to feed itself.

Approximately 9.5 months after initial presentation, the snake was referred to the William R. Pritchard Veterinary Medical Teaching Hospital (Davis, CA) for further diagnostic evaluation and advanced imaging. Physical examination at that time revealed poor condition, with prominent muscle wasting in the cranial and cervical regions. Funduscopic examination was bilaterally normal. Full neurologic examination at rest revealed normal mentation, posture, head carriage, and locomotion. After handling, however, there was cervical weakness exhibited by neck ventroflexion and difficulty in fully lifting the head. The righting reflex was delayed in the cranial one-half of the body. The snake appeared to respond normally to visual cues and room temperature objects on the left, but did not react to these same objects on the right, indicating possible right-sided optical and thermal blindness. Based upon the historical information and the neurologic examination, the neuroanatomical localization was suspected to be a right-sided intracranial lesion.

The snake was consciously intubated with the use of a 2.5-mm uncuffed endotracheal tube and manual intermittent positive pressure ventilation using isoflurane (5%, Forane, Baxter, Deerfield, IL) in 100% oxygen (1 L/min) was administered to achieve a light surgical anesthetic state. After that time, maintenance of anesthesia was achieved with 2% isoflurane with manual ventilation. During anesthesia, heart rate was monitored by Doppler and esophageal stethoscope.

Magnetic resonance imaging (MRI) was performed with the use of a 7-T research unit to obtain sagittal, transverse, and dorsal plane T2-weighted and transverse T1-weighted sequences. There was a well-defined, ovoid (9.1 × 7.0 × 4.5 mm), heterogenous, granular appearing T2W hypointense, T1W isointense mass occupying midline (slightly more right sided) in the dorsal aspect of the midbrain. The mass effaced the tectum and compressed the medulla. The cerebellum could not be identified. Within the mass, there were multiple variably sized foci of marked hypointensity consistent with susceptibility artifact. The ventricular system was dilated and there was asymmetry of the lateral ventricles with the right side larger than the left. (Figs. 1–2). The findings were consistent with a large mass in the dorsal midbrain (suspect extra-axial) with intralesional hemorrhage or mineralization probably causing a mild secondary obstructive hydrocephalus. The primary differentials for such lesion were neoplasia or granuloma.

Based on these findings, a grave prognosis was given and the snake was recovered from anesthesia uneventfully. The owner elected to euthanize the snake but allowed reimaging with a 1.5-T MRI unit, which is more commonly available in a clinical setting. Three days later, the snake was anesthetized with ketamine hydrochloride (30 mg/kg, IM, Ketaset®, Zoetis, Florham Park, NJ). The snake's head was placed in a wrist coil, which was used to obtain sagittal and transverse T1 and T2W sequences as well as a transverse T2*W sequence of the brain. Precontrast findings were consistent with the previous MRI (Fig. 2). A contrast agent (0.2 ml/kg, Magnevist, Bayer Healthcare Pharmaceuticals Inc., Wayne, NJ) was administered intracardiac and a postcontrast transverse T1W sequence was acquired. The previously described large midbrain mass had few areas of ill-defined contrast enhancement and was mildly rim-enhancing. Following completion of imaging, euthanasia was achieved with the use of pentobarbital (0.3 ml/kg, Euthasol ®, Virbac Corp., Fort Worth, TX) followed by potassium chloride (1.3 mEq/kg, Hospira, Boulder, CO) via cardiac injection.

At necropsy there was mild diffuse proteinaceous effusion in the coelomic cavity. Gross examination of the brain revealed a well-circumscribed, bilobulated, tan, and solid firm mass that measured approximately 1 cm in diameter. The mass was located within the brainstem; it distorted and attenuated both diencephalic and mesencephalic regions and obscured the ventricular system at those sites (Fig. 3). All tissues were sampled and immersed in 10% buffered formalin. Subsequently, representative samples of all organs were trimmed, embedded in paraffin, and stained with hematoxylin and eosin. Subgross examination (Fig. 4) confirmed that the mass completely replaced the thalamus and the pineal gland with a bilobar pattern. The mass compressed the ventral portion of the third ventricle and hypothalamus and extended caudally, replacing most of the midbrain with obliteration of the mesencephalic aqueduct. The mass did not invade the striate body, the cerebellum, or medulla oblongata. Histologically, the brain mass was well-circumscribed, densely cellular, and separated into lobules by fibrovascular septa. Each lobule was composed of haphazardly oriented fascicles of neoplastic cells. These fascicles were composed of elongated cells with scant eosinophilic cytoplasm, an oval to fusiform nucleus with dense stippled chromatin, and no apparent nucleoli. Anisocytosis and anisokaryosis were mild, and three mitotic figures were counted in ten high-power (×400) fields. Throughout the tumor, there were multifocal areas of hemorrhage admixed with abundant cholesterol clefts and a few macrophages. Multifocally, there was intratumoral necrosis. Microscopic findings of nonneural organs included hepatic hemosiderosis and lipidosis, and a severe proximal nephron pigmentary nephropathy, moderate glomerulosclerosis with cystic glomerular atrophy, and interstitial fibrosis.

In order to characterize the brain mass, 4-mm sections were processed for immunohistochemistry using primary antibodies specific for glial-fibrillary acidic protein (1:600 dilution, glial fibrillary acidic protein [GFAP], rabbit polyclonal, Dako Z0334, Carpinteria, CA), pan-cytokeratin (1:100 dilution, Lu5, mouse monoclonal, BioCare Medical CM043C, Concord, CA; 1:300 dilution, AE1/AE3, mouse monoclonal, BioGenex MU0110UC, Fremont, CA), S-100 protein (1:600 dilution, rabbit polyclonal, Vector, Burlingame, CA), vimentin (1:300 dilution, 3B4, mouse monoclonal, Dako M7020, Carpinteria, CA), somatostatin (1:1000 dilution, rabbit polyclonal, Immunostar, Hudson, WI), calretinin (1:200 dilution, rabbit polyclonal, Zymed, San Francisco, CA), ionized calcium-binding adapter molecule 1 ([Iba-1], 1:600 dilution, Wako, Richmond, VA), synaptophysin (1:80 dilution, mouse monoclonal, Dako, Carpinteria, CA), periaxin (1:1000, mouse monoclonal, Sigma-Aldrich, St. Louis, MO), and laminin (1:40, goat polyclonal, BioGenex, Fremont, CA). Briefly, following antigen retrieval, slides were rinsed in deionized water and placed in 0.1M phosphate-buffered saline, pH 7.4 (PBS) or TRIS buffered saline, pH 7.6. The antibody diluent was PBS-Tween 20 (0.02%) and blocking reagent were either PBS-Tween 20 (0.2%) and 10% normal horse serum or TBS-Tween 20 (0.025%) and 3% normal goat serum. Sections were blocked for 20 min. After blocking, the primary antibody was applied without rinsing and incubated for 1 h. Antibody-enzymatic binding reactions for AE1/AE3, and for GFAP and calretinin were detected with the Dako Envision System-HRP mouse K4001-Ms Env and rabbit K4003-Rb Env, respectively. Furthermore, Biocare Medical 4Plus Detection System immunolabeling was done for Lu5 and vimentin with secondary biotinylated anti-mouse or for somatostatin and S-100 with secondary biotinylated anti-rabbit antibodies, and subsequently by using a streptavidin complex staining method (4PLUS LL Streptavidin HRP; Biocare Medical). Final detection of any peroxidase immunoreactivity was visualized with NovaRed (Vector SK-4800, Burlingame, CA), following the manufacturer's instructions. Slides were counterstained with Mayer's hematoxylin. Histologic sections containing pertinent normal (nonneoplastic) tissues from snakes were used as laboratory positive controls for each of the above listed antigenic markers. Negative controls consisted of omission of primary antibody and substitution with PBS-Tween 20 (0.02%).

Neoplastic cells were variably immunoreactive to GFAP (Fig. 5), but not to vimentin or cytokeratins. Microglial cells and macrophages, which were immunoreactive to Iba-1, were scattered within the mass and abundant when surrounding the boundaries of the neoplastic mass. A few S-100 immunoreactive cells and neurofilament immunoreactive axons were scattered throughout the mass, but neoplastic cells were not immunoreactive (Fig. 5). There was no intratumoral immunoreactivity to peripheral neural markers (neurofilaments, laminin, and periaxin), whereas peripheral nerves and ganglia present in examined tissues revealed normal labeling (data not shown). Synaptophysin immunohistochemistry was nondiagnostic in this case.

This report describes the full neurologic examination, MRI, and pathologic findings associated with an intracranial space occupying mass in a boid snake. There are many differentials for neurologic signs in boid snakes including traumatic, nutritional, inflammatory, and infectious causes (Bennet and Mehler, 2006). A common reason for neurologic signs in boid snakes is inclusion body disease of boids, recently associated with snake arenavirus (Hetzel et al., 2013). At this time, most neurologic diagnoses in snakes are made at necropsy with the aid of microbiology and histopathology (Bennet and Mehler, 2005). However, this report highlights that with an appropriate diagnostic workup, including advanced imaging, neurologic disease can be diagnosed antemortem. It also emphasizes the importance of correlating premortem diagnostic techniques based on imaging with a complete postmortem examination.

The large intracranial mass described in this snake arose and effaced areas of the diencephalon or caudal forebrain, especially the pineal gland and the thalamus, and effaced areas of the mesencephalon (or midbrain). The main neurologic examination abnormalities noted in this snake were cervical weakness, delayed righting reflex in the cranial half of the body, and suspected right-sided optic and thermal blindness. Disruption of the optic tectum, located in the mesencephalon, would be expected to cause blindness, as was posited in this case. The sensory pit organ of crotaline snakes is innervated by the trigeminal nerve (Newman et al., 1980; Bennett and Mehler, 2005; Moon, 2011), with its afferent nerve fibers terminating in neuronal nuclei located in the rhombencephalon (i.e., pons, medulla and cerebellum) (Newman et al., 1980; Kohl et al., 2014); this is caudal to the lesion described in this case. However, it is likely that pathways for these neurons were disrupted given the large space occupying mass that was present. Cervical weakness, cervical muscle wasting, and delayed righting reflex in the cranial half of the body were likely mediated by accessory nerve neuropathy. In a study in monitor lizards, a distinct accessory nerve was not identified, but the authors postulated that the fibers of this nerve run with the vagus nerve (Barbas-Henry and Lohman, 1984). Unfortunately, all of the cranial nerves and their origination pathways have been published for snakes.

Magnetic resonance imaging generates images with outstanding tissue contrast and good resolution and has been previously performed in reptilian patients. Normal MRI studies have been reported of the head and coelom of juvenile loggerhead sea turtles (Caretta caretta) (Valente et al., 2006; Arencibia et al., 2012), the coelom of red-eared sliders (Trachemys scripta elegans) and yellow-bellied sliders (Trachemys scripta scripta) (Summa et al., 2012), and the brain of garter snakes (Thamnophis sirtalis) (Anderson et al., 2000). In addition, MRI has been used to diagnose neurologic disorders in reptiles, including a Savannah monitor lizard (Varanus exanthematicus) with spinal cord trauma, an aquatic turtle with middle ear abscess (Bennett and Mehler, 2005), and cervical compressive myelopathy in Komodo dragons (Varanus komodoensis) (Zimmerman et al., 2009). In this case, a 7-T magnet provided superior spatial resolution and high signal-to-noise ratio to allow for antemortem diagnosis of an intracranial mass in a small subject. Although the 1.5-T images from this case were not as high quality as the 7-T images, they were considered to be of adequate diagnostic quality and were further enhanced by the addition of a contrast agent, which has the potential to improve contrast sensitivity dramatically. Because of the large size of the mass growing along the rostral brain stem and expanding the diencephalon and mesencephalon, it was difficult to define the location of the mass as extra- or intra-axial through imaging alone. Although surgery was not feasible in this particular case, this case report demonstrates the potential utility of MRI in surgical planning in cases of an operable mass.

The intracranial mass of this snake, arising within the diencephalon and effacing completely the thalamic and pineal regions with secondary compression of the ventral portion of the third ventricle, was ultimately diagnosed as a primary tumor of the central nervous system rather than a granuloma. Based on the location, type of growth, and morphology of the neoplastic elongated cell population arranged in islands or bundles, pinealocytes, astrocytes, ependymal cells, meningothelial cells or peripheral nerve cells were initially included as potential cells of origin for this tumor. Neoplastic cells infiltrated and replaced the neural parenchyma, resulting in a space-occupying mass that compressed adjacent cerebral hemispheres and mesencephalic structures. In domestic animals, meningiomas, ependymomas, pineal tumors, and peripheral nerve sheath tumors (PNSTS) are usually extra-axial tumors with rare invasion into the brain parenchyma (Higgins et al., in press). Some astrocytomas, like pilocytic or gemistocytic sub-types, grow as intra-axial expansile solitary masses that locally infiltrate and replace the neural parenchyma (Higgins et al., in press). Based on immunohistochemistry, neoplastic cells were variably immunoreactive to GFAP, which recognizes intermediate filaments that are present only in astrocytes and ependymal cells (Higgins et al., in press). Immunohistochemistry for S-100 and periaxin, protein markers expressed in schwannomas (and other PNSTs; Higgins et al., in press), did not label neoplastic cells. The pineal gland of snakes is histologically similar to that of domestic animals, containing pinealocytes and astrocytes. In order to confirm the pineal origin of the mass, synaptophysin, somatostatin, and calretinin were pursued in an attempt to identify pinealocytes. Unfortunately, synaptophysin was nondiagnostic since no immunoreactivity was obtained in the control healthy brain; additionally, neoplastic cells were diffusely negative for somatostatin and calretinin immunostaining. In conclusion, based upon the infiltrative nature of this intraxial mass, and the immunoreactivity of the neoplastic cells to GFAP, the tumor in this animal was consistent with a mass of astroglial or ependymal origin. It is uncertain why astrocytes in the normal, unaffected portion of the brain in this animal were immunoreactive to vimentin, but the neoplastic, GFAP-positive cells lacked vimentin immunoreactivity.

Other differentials for the intracranial mass of this patient included initially a granuloma, abscess, extraneural neoplasia, and xanthomatosis (Bennett and Mehler, 2005). However, these differentials were excluded based on the lack of characteristic features on routine HE sections, and the lack of macrophages and immunoreactivity of intratumoral cells to the macrophage marker Iba-1. Similarly, histopathological examination of the mass excluded a diagnosis of a pituitary tumor such as adenomas or cystadenomas, which are the most common neoplastic intracranial masses reported in snakes (Gyimesi and Garner 2007; Dadone et al., 2010). According to these reports, pituitary tumors do not tend to invade adjacent neural parenchymal structures.

This case highlights the diagnostic utility of MRI in antemortem diagnosis of intracranial masses of snakes. In addition, although inclusion body disease of snakes is a common cause of neurologic signs in boid species, other differentials exist and can occur. As our captive populations of snakes and other reptiles are afforded better husbandry parameters, these animals will be living longer lives and other important differentials, including neoplasia, should be considered as causes for decline in health or condition.