Abstract
Case history
A 12.5-year-old 6.5 kg castrated male domestic shorthair cat was presented to the Iowa State University Veterinary Medical Center for evaluation of acute respiratory distress. He had a 3-week history of coughing and sneezing for which he had been treated with an unspecified course of corticosteroids. At the time of presentation, he was receiving amoxicillin-clavulanic acid (19 mg/kg PO q12h) and enrofloxacin (3.5 mg/kg PO q24h). His clinical signs improved briefly after the initial steroid injection, but signs worsened acutely the night of presentation. He was an indoor cat with an up-to-date vaccination record.
Physical examination and preliminary investigations
On physical examination, the cat was tachypneic (60 breaths per min) with an increased respiratory effort and prominent referred upper airway sounds. No murmur or arrhythmia was auscultated. The remainder of the physical examination was unremarkable. Thoracic radiographs were within normal limits.
Minor changes were present on hematology and serum biochemistry. A complete blood count revealed lymphopenia (1.25 × 10 3 /μl; reference interval [RI] 1.5–7.0 × 10 3 /μl) and eosinophilia (0.94 × 10 3 /μl; RI 0.0–0.75 × 10 3 /μl). Mild thrombocytopenia was noted (273 × 10 3 /μl; RI 300–800 × 10 3 /μl); however, platelets were clumped and appeared adequate on a blood smear. Serum biochemical analysis revealed the following abnormalities: hyponatremia (149 mEq/l; RI 155–165 mEq/l), hypochloridemia (106 mEq/l; RI 123–131 mEq/l) and elevated bicarbonate (26.8 mEq/l; RI 17–24 mEq/l). Albumin was slightly elevated at 4.1 g/dl (RI 2.1–3.5 g/dl), consistent with dehydration. There were also slight elevations in alanine aminotransferase, at 184 IU/l (RI 20–125 IU/l), and anion gap, at 20 (RI 12–16). Serology results for feline leukemia virus (FeLV), feline immunodeficiency virus (FIV) and heartworm were negative.
Urine was obtained via cystocentesis and submitted for urinalysis. Urine specific gravity was 1.040. Hematuria was noted (blood 3+, red cells 30–50/high power field) along with 2+ bacteria. Bacterial culture was negative.
On repeat physical examination later that day, there was severely diminished nasal airflow bilaterally and stertorous respiration. The primary clinical problem was upper respiratory disease. Differential diagnoses included neoplasia and inflammatory or infectious rhinitis/nasopharyngitis (bacterial, fungal or viral).
Computed tomography and rhinoscopy
Computed tomography (CT) of the skull was performed, which revealed a contrast-enhancing 1 cm × 0.5 cm soft tissue mass filling the middle portion of the nasopharynx (Fig 1). The mass was centered near the caudal aspect of the hard palate, extending 1 cm craniad and 3 mm caudad to that point. The caudal nasopharynx was open and clear at the level of the caudal soft palate. The lymph nodes of the head were normal and there was no evidence of invasion of the cribriform plate.
Post-contrast computed tomography axial view, showing the nasopharyngeal mass (arrow)
Rhinoscopy was performed following the CT scan. The oropharynx was normal. The soft palate was retracted and the bronchoscope (Olympus BF-P160, light source CVL 160) was retroflexed into the nasopharynx. A 1 × 0.5 cm pink fleshy soft tissue mass was visualized (Fig 2). Several biopsies were taken and submitted for histologic assessment.
This image of the nasopharyngeal mass was obtained via retroflexion of a bronchoscope dorsally and cranially behind the soft palate. The vomer bone is seen dorsally, the soft palate is seen ventrally and the nasopharynx is seen centrally
Histology
Histopathology of the mass revealed an invasive population of polygonal cells forming nests separated by moderate amounts of an arborizing fibrovascular stroma. Neoplastic cells were arranged in peritrabecular palisades, Homer-Wright rosettes and Flexner-Wintersteiner rosettes (Fig 3). Neoplastic cells had basal, hyperchromatic nuclei and moderate amounts of eosinophilic cytoplasm that tapered toward the rosette's center. Nuclei were moderately sized, round to irregular, hyperchromatic and had multiple small nucleoli. Anisocytosis and anisokaryosis were minimal and mitotic figures were not observed.
Histopathology of the olfactory neuroblastoma showing peritrabecular palisades (arrow-head), HomerWright rosettes (arrows) and (inset picture) a Flexner-Wintersteiner rosette (arrow-head). Magnification 600x
On immunohistochemical testing, the tumor stained positive for keratin (Fig 4), neuron-specific enolase and vimentin, and negative for glial fibrillary acid protein, neurofilaments and S-100.
Immunohistochemistry. Keratin stain. Magnification 600x
The final diagnosis was olfactory neuroblastoma (esthesioneuroblastoma).
Outcome
Treatment options were discussed with the owner, including radiation therapy and chemotherapy. Due to financial constraints, the owner elected to take the cat home. Two weeks later, the owner reported that the cat was still doing well, although he was snorting occasionally. A further 8 days later, the cat presented to his local veterinarian for recurrence of stertorous, open-mouth breathing. At that point, the owner elected humane euthanasia. No necropsy was performed.
Discussion
Olfactory neuroblastoma is a rare malignant tumor of the olfactory neuroepithelium. It was first reported in human medicine in 1924. 1 It is widely accepted that olfactory neuroblastoma originates from the olfactory membrane of the upper nasal cavity. 2 It has also been reported to invade the intracranial cavity via the cribriform plate. 3 Reports of this neoplasm are infrequent in the veterinary literature, but cases have been described in several species. 4
Tumor frequency and clinical signs
Nasal and paranasal sinus tumors have been reported to represent 1–4.2% of all feline neoplasms. The most common feline nasal neoplasms include adenocarcinoma, squamous cell carcinoma and lymphoma, although there is controversy over their relative prevalence. 5–7 Over 90% of these tumors are malignant. 5,7,8 Historically, older castrated male cats have been predisposed to nasal tumors. 5,6,8 Given the low numbers of reported feline olfactory neuroblastomas, it is difficult to say if the same criteria apply.
Clinical signs of olfactory neuroblastoma are non-specific. In human patients, the most common presenting complaints are progressive nasal obstruction, epistaxis and hypo/anosmia. In some instances, there are no clinical signs noted at the time of diagnosis. 9 Veterinary patients with nasal and paranasal sinus tumors commonly present with sneezing, nasal discharge, dyspnea, facial deformity, stertor and partial to complete anorexia. 5,7,8 Alternatively, patients may present for evaluation of seizures and/or neurologic signs. 3,10
Histologic diagnosis
Diagnosis of olfactory neuroblastoma relies on certain histologic criteria. The tumor must exhibit lobular architecture with a vascular interlobular stroma. Instead of having distinct cytoplasmic borders, the nuclei are surrounded by neurofibrillary material, often forming rosette patterns. A grading system based on histopathology exists — grade I representing the most differentiated tumors, and grade IV the least differentiated. Grade I and II tumors frequently exhibit a pseudorosette pattern (Homer-Wright type), whereas grade III tumors tend to demonstrate true neural rosettes (Flexner-Wintersteiner type). Neither type of rosette is typically seen in grade IV tumors. Grade IV neoplasms are the most difficult to diagnose and are often confused with small cell undifferentiated tumors of the sinonasal tract. Grade III and IV tumors may require further ultrastructural, immuno-peroxidase or fume-fluorescence studies to distinguish them from poorly differentiated carcinomas. Immunohistochemistry stains are routinely performed to aid in the diagnosis of olfactory neuroblastoma. 2
In human medicine, there is a pre-therapy staging system for olfactory neuroblastoma. 11,12 Tumors are staged into one of four groups based on anatomic location. Staging systems for canine nasosinal tumors have been described, focusing on primary tumor location and extent of tumor involvement. 13 Most staging systems have not been proven to have prognostic significance. Feline-specific systems have not been described.
Treatment protocols and prognostic factors
Treatment protocols and success rates have been infrequently reported in the veterinary literature. The only report the authors managed to locate describes successful radiotherapy treatment of a dog with olfactory neuroblastoma. 14 Given its rarity in human medicine (olfactory neuroblastoma represents 3% of all intranasal neoplasms), the ‘best’ treatment has yet to be established. There seems to be a consensus that aggressive surgical debulking (ie, craniofacial resection) followed by a combination of radiotherapy ± chemotherapy has offered the best long-term survival rates. 9,15 Most often, the tumor stage and grade will influence the treatment protocol. Advanced stage tumors are often not amenable to surgery. In these cases, radiation therapy, chemotherapy or chemoradiotherapy is recommended. 15
One human study evaluated whether the expression of tumor markers in olfactory neuroblastomas could predict the response to chemotherapy and the prognosis. The authors concluded that the tumor marker bcl-2, a member of the gene family that regulates apoptosis, was prevalent in olfactory neuroblastoma and that expression correlated with a good initial response to chemotherapy. However, studies have demonstrated that overall survival tends to be poorer in patients with tumors expressing bcl-2. 16
A few veterinary studies have attempted to characterize the expression of bcl-2 in healthy and neoplastic feline tissue. In one study, bcl-2 was consistently expressed in the basal layers of mucous membrane epithelium in healthy cats. The same study identified bcl-2 expression in intranasal carcinomas. 17 Another study evaluated the expression of bcl-2 in cats with lymphoma. No correlation was found between bcl-2 expression and response to chemotherapy. 18 Feline olfactory neuroblastomas have not been evaluated for bcl-2 expression.
In one veterinary report, numerous type C FeLV retroviral particles were identified via polymerase chain reaction in olfactory neuroblastomas diagnosed in three cats. The retroviral particles were similar in appearance to the neurosecretory granules that are highly characteristic of olfactory neuroblastoma in humans. The authors suggest that there may be a relationship between FeLV and the development of olfactory neuroblastoma, and that FeLV particles may have been misdiagnosed in other feline olfactory neuroblastoma cases as neurosecretory granules. 19 Further investigation is warranted to evaluate the presence of retroviral particles in feline olfactory neuroblastoma and whether a true causal relationship exists.
Tumor recurrence is the primary cause of treatment failure is most patients. Several authors agree that higher stage and grade tumors are more likely to recur. Olfactory neuroblastoma most commonly metastasizes to regional lymph nodes. The most commonly affected distant sites are lung, brain and bone. 9,15,20,21
Conclusions
Olfactory neuroblastoma is a differential for spontaneous intranasal tumors in cats. Clinical signs may be variable and depend on the extent of tumor involvement. If histologic changes are not definitive, special stains may be required to correctly identify the nasal tumor as an olfactory neuroblastoma. Treatment options are varied and somewhat dependent upon surgical skill and the availability of adjuvant chemoradiotherapy.