Use of computed tomography imaging during long-term follow-up of nine feline tuberculosis cases

Abstract

Case series summary

Feline tuberculosis is an increasingly recognised potential zoonosis of cats. Treatment is challenging and prognosis can vary greatly between cases. Pulmonary infection requires extended courses of antibiotics, but methodologies for sensitively monitoring response to treatment are currently lacking. In this case series, we retrospectively examined the serial computed tomography (CT) findings in nine cats that had been diagnosed with tuberculosis. Changes in pathology (where applicable to tuberculosis) were correlated with the clinical presentation of each of the cats, the treatment protocol, and previous and contemporary diagnostic investigations. This study found that changes in CT findings during the medium- to long-term management of feline tuberculosis were highly variable between cats. The majority of cats had reduced pathology at re-examination during anti-tuberculous therapy, but pathology only resolved in a minority of cases. In some cases recurrence of pathology detected by CT imaging preceded clinical deterioration, allowing for rapid therapeutic intervention.

Relevance and novel information

When considered in combination with clinical findings, CT studies can aid in decision making regarding tapering of antibiotic protocols, or reintroduction of therapy in cases of recurrence or reinfection. This series also highlights that, in some cases, persistent abnormalities can be detected by CT, so complete resolution of CT pathology should not always be a goal in the management of feline tuberculosis.

Introduction

Feline tuberculosis is a highly variable and increasingly recognised disease in domestic pet cats in the British Isles.^1–3 Infection is assumed to be acquired from bites by prey species sustained during hunting, leading to the most typical clinical presentation of cutaneous lesion(s) at ‘fight and bite sites’ with or without regional lymph node involvement.^1–3 Disseminated disease can occur, resulting in non-specific signs related to the respiratory and/or alimentary tracts, giving rise to variable findings on diagnostic imaging investigations.^4–7 Thoracic and/or abdominal pathology can more rarely result from acquisition of disease through inhalation or ingestion.^1,5 The radiological and computed tomography (CT) abnormalities associated with disseminated mycobacterial infection have previously been described.^2,4,7

Advocated treatment protocols for feline tuberculosis typically consist of an initial and a continuation phase. ⁸ The initial phase combines three antibiotic drugs lasting for 2 months, while the continuation phase comprises two drugs for a further 4 months. ⁸ However, it is possible that treating with all three drugs until 2 months after apparent clinical resolution, which typically results in 4–6 months of treatment, may result in a better clinical outcome (DGM and COH, unpublished data, 2016).

Prognosis varies depending on the species of mycobacterium involved, the extent and severity of disease, and the compliance and tolerance of the patient to medication.^1,6 While many cases respond favourably to therapy, resulting in apparent cure or long-term remission, other patients either fail to respond or go on to develop recurrence of signs following apparently successful treatment.^1,6

In order to assist clinical decision making by veterinary surgeons and owners, a reliable method is needed to monitor the disease at all stages of management. The use of CT has already been shown to be a valuable tool in the initial diagnosis. ⁷ In this report, we describe the use of CT during the medium- and long-term follow-up of tuberculous disease in nine cats between June 2010 and May 2016. Table 1 shows signalment and summary data for all nine cases detailed below.

Table 1.

Summary details of the nine case of feline tuberculosis where serial CT images were used as part of clinical follow-up

Case number	Breed	Age (years)	Gender	Location in UK	Weight (kg) at initial presentation	Haematology and serum biochemistry (reference interval)	FIV / FeLV status	Diagnosis	Impact of CT evaluations
1	Oriental	7	MN	South Scotland	5	Total calcium 3.13 mmol/l (1.95–2.83 mmol/l) Ionised calcium 1.75 mmol/l (1.05–1.45 mmol/l)	Negative	M microti	Early reinstigation of antibiotic therapy following slight clinical deterioration
2	DSH	11	FN	Central Scotland	3.6	No abnormalities detected	Negative	M microti	Pulmonary dissemination of tuberculosis diagnosed. Mid-term static appearance of lesion irrespective of antibiotic therapy
3	Bengal	13	MN	South Scotland	5	No abnormalities detected	Negative	M microti	Delayed antibiotic tapering due to persistent abnormalities. Early reinstigation of antibiotic therapy following slight clinical deterioration
4	British Shorthair	10	MN	Cheshire, England	3.8	Hyperglobulinaemia	Negative	M bovis	Reduction of antibiotic therapy with improvement in detectable abnormalities
5	DSH	7 months	MN	Bristol, England	2.8	Total calcium 3.95 mmol/l (2.30–2.50 mmol/l)	Negative	M microti	Discontinuation of antibiotic therapy with improvement in detectable abnormalities
6	DSH	3	FN	West Midlands, England	4.1	No abnormalities detected	Negative	Tuberculosis complex	Reduction of antibiotic therapy with early improvement in detectable abnormalities
7	DSH	7	MN	South Scotland	5	No abnormalities detected	Negative	M microti	Reduction of antibiotic therapy with improvement in detectable abnormalities
8	Burmilla	8	ME	South Scotland	4.6	No abnormalities detected	Negative	M microti	Discontinuation of antibiotic therapy with improvement in detectable abnormalities
9	DSH	7	MN	Central Scotland	5.7	No abnormalities detected	Negative	M microti	Continuation of antibiotic therapy with partial improvement in detectable abnormalities

DSH = domestic shorthair; MN = male neutered, FN = female neutered; ME = male entire; FIV = feline immunodeficiency virus; FeLV = feline leukaemia virus; M microti = Mycobacterium microti; M bovis = Mycobacterium bovis

Case series description

Case 1

Case 1 initially presented with anorexia and weight loss. Mild mandibular lymphadenomegaly and harsh lung sounds were noted on physical examination. Thoracic radiographs revealed a diffuse structured interstitial lung pattern; CT was not performed as the clinic did not have on-site access to CT at this time. The feline interferon gamma (IFN-γ) release assay (IGRA) was performed by Biobest Laboratories, Edinburgh, and indicated infection with Mycobacterium microti. ⁸ The cat was treated with a triple antibiotic protocol of rifampicin (generic; Mylan) (10 mg/kg) 50 mg PO q24h, marbofloxacin (Marfloquin; Virbac) (3 mg/kg) 15 mg PO q24h, and azithromycin (Zithromax; Pfizer) (6 mg/kg) 30 mg PO q24h for 2 months as the induction treatment phase; marbofloxacin was then discontinued and the remaining antibiotics continued for the maintenance phase. After 6 months, clinical remission from disease was achieved; serum calcium concentration was within the reference interval and repeat radiographs revealed no abnormalities, so antibiotic therapy was stopped.

Eleven months after antibiotic treatment had been discontinued, the cat represented with a recurrence of lethargy and anorexia, with normal lung sounds but reduced thoracic compression. Body weight had increased to 6.2 kg. A recurrence of hypercalcaemia was noted (ionised calcium 1.75 mmol/l) and serum 25-hydroxyvitamin D concentration was low (46 pg/ml, RI 14.9–61.0 ng/ml). Full-body CT was performed using a VetMouseTrap device, revealing mild tracheobronchial, mediastinal and mesenteric lymphadenomegaly and a diffuse, moderate reticulonodular lung pattern (Figure 1a). Recurrence or reinfection of tuberculosis was assumed and triple antibiotic therapy was reinstated (drugs and doses as above, dosed for a 6 kg cat). In addition, calcitriol supplementation was given at a dose of 2 μg/kg PO q24h. Three months later the cat was reassessed, and clinical examination and whole-body CT were normal (Figure 1b). On the basis of completing 3 months of triple antibiotic therapy and resolution of clinical signs, treatment was changed to pradofloxacin (Veraflox tablets; Bayer) (4 mg/kg) 25 mg PO q24h, which was given as an antimicrobial monotherapy for 6 months with calcitriol supplementation as previously described. Two further CT examinations were performed, at 4 and 6 months after disease recurrence, and were normal.

Figure 1.

CT appearance of lung parenchyma in case 1 at the level of the accessory lung lobe on three different occasions. (a) Diffuse, moderate reticulonodular pattern identified on the first occasion of disease recurrence following 11 months of clinical remission. (b) Normal pulmonary appearance 3 months later following triple antibiotic therapy and calcitriol supplementation. (c) Diffuse, mild reticular pattern noted concurrent with an increased sleeping respiratory rate but normal clinical examination, indicative of probable tuberculosis recurrence/relapse 11 months after image (a)

Eleven months after recurrence, after 2 months off pradofloxacin, the cat was represented as the owner observed a mildly increased sleeping respiratory rate (21 bpm; this cat’s normal sleeping respiratory rate was <20 bpm). Despite a normal clinical examination, a CT scan demonstrated a diffuse mild reticular lung pattern with areas of ground glass opacity (Figure 1c); the serum calcium concentration was increased and serum 25-hydroxyvitamin D concentration was low. Triple antibiotic therapy was restarted (rifampicin and azithromycin, dosed as above, plus pradofloxacin [Veraflox liquid; Bayer] [~5 mg/kg] 30 mg PO q24h), and calcitriol treatment was restarted at (2 μg/kg) 12.5 μg PO q24h (body weight 6.5 kg). After 2 months of treatment, repeat CT examination was normal. Due to the history of several episodes of disease, it was recommended that the triple antibiotic therapy be continued for a further 4 months, followed by 3 months of double antibiotic therapy (azithromycin and pradofloxacin, dosed as above). The cat remained clinically normal throughout this period and treatment was discontinued a total of 20 months after the initial recurrence. Two months later, another IGRA returned a negative result and the serum calcium and 25-hydroxyvitamin D concentrations were within normal limits.

A further episode of mycobacterial recurrence/reinfection occurred after 8 months without treatment. The cat was again re-presented following observation of a mildly increased sleeping respiratory rate (23 bpm; body weight 7.1 kg). Whole body CT demonstrated mild diffuse thoracic and abdominal lymphadenomegaly, and a diffuse but patchy, mild-to-moderate reticulonodular lung pattern. A repeated IGRA was positive and consistent with M microti infection. Triple antibiotic therapy was prescribed for 3 months (rifampicin, pradofloxacin and azithromycin, dosed as above, for a 7 kg cat), followed by double antibiotic therapy for a further 9 months (pradofloxacin and azithromycin, dosed as immediately above). During this period, the cat remained well, and a further four full-body CT examinations revealed a normal pulmonary parenchymal appearance. Given the normal imaging and clinical findings throughout this period, antibiotics were discontinued as planned, and the cat remains well without recurrence of clinical signs over 17 months later; during this time 5 CT scans revealed no detectable abnormalities. A timeline of this case is shown in Figure 2.

Figure 2.

Timeline of diagnostic investigations and treatment for case 1; a 7-year-old male neutered Oriental cat with pulmonary tuberculosis caused by Mycobacterium microti. Rad = radiograph; TB = tuberculous changes; NAD = no abnormalities detected; mn = months; T = treatment; R = rifampicin; A = azithromycin; M = marbofloxacin; V = vitamin D; P = pradofloxacin; TB? = potentially tuberculous changes

Case 2

Case 2 was first presented for weight loss and generalised lymphadenomegaly. Radiographs revealed a diffuse interstitial lung pattern (CT was not available at the clinic at that time). Excisional biopsy of the popliteal lymph nodes was performed; histopathology revealed a granulomatous lymphadenitis and Zeihl Neelsen (ZN) staining identified intralesional acid-fast bacilli indicative of mycobacterial infection. A triple antibiotic protocol was instigated (rifampicin [11 mg/kg] 40 mg PO q24h; marbofloxacin [2.7 mg/kg] 10 mg PO q24h; clarithromycin [11 mg/kg] 40 mg PO q12h) for 2 months followed by rifampicin and marbofloxacin (same doses) for 4 months. Revisits revealed initially static peripheral lymphadenomegaly, which resolved over the 4 months of maintenance treatment. Repeat thoracic radiography at the end of the maintenance phase revealed no abnormalities and treatment was therefore discontinued.

Four months following the end of treatment the cat presented to an emergency clinic with acute respiratory signs. Laryngeal swelling was identified and following stabilisation with corticosteroids, furosemide, chlorphenamine (all at standard doses), plus additional oxygen, the laryngeal swelling resolved. Radiography revealed a thoracic mass consistent with an enlarged cranial mediastinal lymph node. This was confirmed on full body CT examination using a VetMouseTrap device, which also revealed moderate mineralisation within the mass lesion (Figure 3a). Fine needle aspiration (FNA) of the mass yielded a non-diagnostic sample, while an IGRA was consistent with M microti infection. Given the previous history of mycobacterial lymphadenitis, with an owner who was reticent to restart triple therapy, the cat was started on single antibiotic therapy (pradofloxacin liquid [7 mg/kg] 25 mg PO q24h) to see if this might reduce the size of the thoracic mass and so give weight to the diagnosis that it may be tuberculous. One month later, the cat was clinically well and CT revealed a static appearance to the mass. Antibiotic therapy was discontinued as it did not appear to be effective. Three months later the CT appearance remained unchanged, and a repeat IGRA was inconclusive.

Figure 3.

CT images at the level of the third sternebra from case 2 on two different occasions. (a) Image acquired 4 months after cessation of antibiotic therapy for disseminated tuberculosis, showing an enlarged cranial mediastinal lymph node (*). (b) Image acquired 5 months later, showing a static appearance of the lymph node but a mild ground-glass appearance of the adjacent lung parenchyma (arrow) indicative of regional extension of disease. The cat was concurrently diagnosed with a mandibular squamous cell carcinoma

The cat re-presented the next month with hypersalivation and difficulty eating. Physical examination revealed thickening of the caudal aspect of the right mandibular ramus, with loosening of the associated teeth. On CT, this lesion was characterised by moderate bone lysis with concurrent proliferation; moderate regional lymphadenomegaly was noted. The thoracic mass remained static in appearance, but the surrounding lung had a mild patchy ground-glass appearance (Figure 3b). The appearance of the mandibular lesion was not considered typical for tuberculous osteomyelitis. Biopsy of the mandibular mass and local lymph nodes resulted in a diagnosis of squamous cell carcinoma with reactive lymphoid hyperplasia. The owner opted for palliative therapy with meloxicam (Metacam; Boehringer Ingelheim, 0.05 mg/kg PO q24h), and after 3 weeks the cat was euthanased. Post-mortem examination was performed and histopathology of the enlarged cranial mediastinal lymph node revealed large numbers of acid-fast bacilli within the node and the peri-nodal connective tissue. As indicated by CT, granulomatous inflammatory changes extended into the adjacent pulmonary parenchyma. The lymph node was confirmed to be PCR positive for M microti by the Mycobacterial Reference Laboratory, Leeds University Teaching Hospital. A timeline of this case is shown in Figure 4.

Figure 4.

Timeline of diagnostic investigations and treatments for cases 2–9. Rad = radiograph; US = ultrasound; TB = tuberculous changes; NAD = no abnormalities detected; mn = months; T = treatment; R = rifampicin; A = azithromycin; M = marbofloxacin; C = clarithromycin; V = vitamin D; P = pradofloxacin; TB? = potentially tuberculous changes; Euth = euthanasia; SCC = squamous cell carcinoma; No = no treatment given; Sx = surgery; MN = male neutered; FN = female neutered; y = year-old; DSH = domestic shorthair; BSH = British Shorthair; M microti = Mycobacterium microti

Case 3

Case 3 initially presented with mandibular lymphadenomegaly. Sternal lymphadenomegaly was noted on thoracic radiography, and abdominal ultrasound revealed marked mesenteric lymphadenomegaly and focal marked circumferential jejunal thickening; FNA of the mandibular and jejunal lymph nodes and the abnormal jejunal wall revealed granulomatous inflammation with acid-fast bacilli indicative of mycobacterial infection. An IGRA was consistent with M microti infection and the cat was started on triple antibiotic therapy (rifampicin [10 mg/kg] 50 mg PO q24h; azithromycin [8 mg/kg] 40 mg PO q24h; pradofloxacin tablets [5 mg/kg] 25 mg PO q24h); and calcitriol supplementation ([2 μg/kg] 10 μg PO q24h). Two months later the cat was clinically well, although the right mandibular lymph node remained slightly enlarged. A conscious full-body CT examination using a VetMouseTrap device was performed, revealing improved but persistent mesenteric lymphadenomegaly. Given the clinical and imaging findings, the triple antibiotic therapy described above was maintained for another 4 months, giving a total treatment duration of 6 months, after which the mandibular and mesenteric lymph nodes were palpably normal and antibiotic therapy was discontinued (body weight 6.4 kg at this time).

Three months later the cat re-presented with weight loss, lethargy and inappetence (body weight 6.0 kg). The peripheral lymph nodes were of normal size, but harsh inspiratory lung sounds and multiple palpable abdominal masses were noted. Both abdominal ultrasound and full-body CT were performed, confirming the presence of marked thoracic and abdominal lymphadenomegaly, and focal marked jejunal thickening as has been previously described. A diffuse, mild reticulonodular lung pattern was also noted. A FNA of the mesenteric lymph nodes again revealed granulomatous inflammation with acid-fast bacilli. Triple antibiotic therapy was resumed at the dose rates detailed previously, but despite an initially improved demeanour the cat continued to lose weight and after 5 months of treatment was euthanased. Post-mortem examination was not performed. A timeline of this case is shown in Figure 4.

Case 4

Case 4 initially presented with weight loss, dyspnoea and coughing. Physical examination revealed tachypnoea (respiratory rate 40 bpm), with increased inspiratory and expiratory effort and noise. Thoracic CT examination revealed a moderate multifocal alveolar pattern with regions of pulmonary cavitation affecting multiple lung lobes, most marked within the right caudal lobe, and a moderate thoracic lymphadenomegaly (Figure 5a). A right caudal lung lobectomy was performed and histopathology revealed necrotising and pyogranulomatous bronchopneumonia; however, no acid-fast bacteria were identified. Tissue was submitted for culture and blood for IGRA, and treatment with marbofloxacin ([2 mg/kg] 8 mg PO q24h) was started. A good clinical response was noted in the initial 2 month postoperative period; however, tissue culture and IGRA both confirmed Mycobacterium bovis infection, and a standard triple antibiotic protocol was introduced (marbofloxacin [2 mg/kg] 8 mg PO q24h; azithromycin [10 mg/kg] 40 mg PO q24h; rifampicin [20 mg/kg] 80 mg PO q24h – although the dose of rifampicin was high).

Figure 5.

CT appearance of the lung parenchyma in case 4 at the level of the accessory lung lobe on two different occasions. (a) Multifocal regions of alveolar pattern with associated pulmonary cavitation (*) identified at initial presentation. (b) Follow-up imaging after right caudal lung lobectomy and 8 months of antibiotic treatment shows residual patchy ground-glass appearance and collapsed pulmonary cavities (arrow). An additional CT study performed 4 months postsurgery (not shown) showed very similar residual changes

After 2 months of triple antibiotic treatment, CT was repeated revealing residual patchy ground-glass opacity, with collapsed cavities within the remaining lung lobes, but subjectively normal thoracic lymph nodes. Due to the improved pulmonary appearance and the good clinical condition of the cat, triple antibiotic therapy was reduced to dual therapy (marbofloxacin and rifampicin, dosed as above). After a further 4 months, the appearance of the lungs on CT examination was unchanged (Figure 5b) and a repeat IGRA remained positive. Antibiotic treatment was discontinued, and the cat remained well, with a negative IGRA result obtained 6 months later. A timeline of this case is shown in Figure 4.

Case 5

Case 5 initially presented with coughing, resting tachypnoea (respiratory rate 55 bpm) and exercise intolerance. Body weight and condition score (1.5/5) were low. Thoracic and abdominal CT examination revealed a diffuse marked nodular lung pattern with occasional scattered foci of pulmonary mineralisation (Figure 6a), marked tracheobronchial lymphadenomegaly and mild peripheral and abdominal lymphadenomegaly. A FNA of lung tissue revealed marked pyogranulomatous inflammation with acid-fast bacilli and was PCR positive for M tuberculosis complex organisms. The IGRA suggested infection with M microti. A standard antibiotic protocol of 2 months’ triple therapy (pradofloxacin [~5 mg/kg] 15 mg PO q24h; azithromycin [~10 mg/kg] 30 mg PO q24h; rifampicin [~10 mg/kg] 30 mg PO q24h) was followed by ongoing double therapy (azithromycin and rifampicin, dosed as above).

Figure 6.

CT appearance of the lung parenchyma in case 5 at the level of the accessory lung lobe on two different occasions. (a) Marked, diffuse nodular lung pattern with occasional foci of mineralisation (arrows) identified at initial presentation. (b) Follow-up imaging after 8 months of treatment shows a persistent mild reticulonodular pattern with mildly more extensive parenchymal mineralisation (arrow). Treatment was discontinued and a static appearance was recorded 12 months later, indicating these persistent changes do not reflect active disease

At a recheck after 8 months of treatment the cat was clinically normal and had an improved body weight and body condition score (4.4 kg and 2.5/5). Thoracic CT revealed only a mild diffuse reticulonodular lung pattern, but scattered pulmonary mineralisation was more extensive than previously noted (Figure 6b). Antibiotic therapy was discontinued. The cat remained well and the CT abnormalities were seen to be static at a revisit 12 months later. A timeline of this case is shown in Figure 4.

Case 6

Case 6 presented with lethargy, intermittent dyspnoea, weight loss, stridor and nasal discharge. Clinical examination revealed a moderate inspiratory dyspnoea with wheezing on auscultation, bilateral serous nasal discharge, bilateral renomegaly and bilateral popliteal lymphadenomegaly. A CT examination of the head, thorax and abdomen revealed an alveolar lung pattern within the right middle and ventral right caudal lung lobes, with a diffuse moderate reticulonodular pattern, moderate multifocal lymphadenomegaly, mild bone lysis over the nasal bridge and multiple mass lesions in both kidneys. Nasal biopsies confirmed mycobacterial infection by histopathology, and was PCR positive for M tuberculosis complex organisms, but the laboratory was unable to further define the species. A standard antibiotic protocol of 2 months’ triple therapy was prescribed (pradofloxacin [~5 mg/kg] 20 mg PO q24h; azithromycin [~10 mg/kg] 40 mg PO q24h; rifampicin [~10 mg/kg] 40 mg PO q24h), followed by ongoing double therapy (pradofloxacin and rifampicin, dosed as above).

Two months after the start of antibiotic therapy the cat was clinically well. The CT showed marked improvements, with residual diffuse mild pulmonary ground-glass appearance, mild multifocal lymphadenomegaly and partial resolution of the renal mass lesions. Antibiotics were discontinued after a 6 month course, and the cat remained clinically well 12 months later. A timeline of this case is shown in Figure 4.

Case 7

Case 7 presented with dysuria due to a well-demarcated alopecic skin nodule of 2 cm diameter over its prepuce. Physical examination revealed a mildly elevated resting respiratory rate (48 bpm). An incisional biopsy of the preputial lesion revealed granulomatous inflammation and rare acid-fast bacilli indicative of mycobacterial infection. An IGRA was strongly suggestive of an M microti infection. A CT scan, performed using a VetMouseTrap device, revealed a focal region of alveolar pattern in the left cranial lung lobe with a diffuse mild reticulonodular pattern suggestive of pulmonary tuberculosis. The cat was placed on standard triple antibiotic therapy (pradofloxacin tablets [3 mg/kg] 15 mg PO q24h; azithromycin [6 mg/kg] 30 mg PO q24h; rifampicin [10 mg/kg] 50 mg PO q24h) for 4 months. By re-evaluation, the preputial lesion and dysuria had completely resolved and thoracic CT revealed an improvement in both the focal and diffuse pulmonary changes. The cat was changed to dual antibiotic therapy (rifampicin and azithromycin, dosed as above), and this was discontinued after an additional 2 months; the cat remains clinically well 6 months later. A timeline of this case is shown in Figure 4.

Case 8

Case 8 was presented for investigation of dyspnoea (respiratory rate 60 bpm), bilateral mandibular lymphadenomegaly and palpable abdominal masses. Abdominal ultrasound showed a diffusely heterogeneous appearance to the spleen and mild generalised abdominal lymphadenomegaly. An exploratory laparotomy was performed to biopsy the abnormal structures. Histopathological analysis of the spleen and medial iliac lymph node revealed granulomatous splenitis and reactive lymphoid hyperplasia consistent with mycobacteriosis, although no acid-fast bacteria were seen. Thoracic radiography revealed a severe diffuse mixed bronchial and nodular pattern with multiple foci of mineralisation in the caudodorsal lung fields. No thoracic lymphadenomegaly was evident. An IGRA indicated M microti infection, so triple antibiotic therapy was instigated for 6 months (marbofloxacin [2 mg/kg] 10 mg PO q24h; rifampicin [16 mg/kg] 75 mg PO q24h; clarithromycin [8 mg/kg] 35 mg PO q12h).

Re-evaluation after 6 months revealed that the initial clinical signs had resolved, and a full body CT scan using the VetMouseTrap identified complete resolution of the previously noted lung pattern and abdominal lymphadenomegaly. Several small mineral foci remained visible within the lungs that were predominantly, but not exclusively, airway associated. Antibiotic therapy was discontinued at this point. The cat remained clinically well and at a routine revisit over 33 months later a full body CT was repeated using the VetMouseTrap. This study revealed normal pulmonary parenchyma and there was no evidence of lymphadenomegaly. More extensive and more widely distributed predominantly airway-associated mineralisation was present. A timeline of this case is shown in Figure 4.

Case 9

Case 9 was presented for investigations into stertorous breathing and a rapidly growing inter-ocular skin lesion. The CT examination of the head and thorax revealed a soft tissue mass lesion overlying the frontal and nasal bones with several associated small foci of bone lysis, and a diffuse but asymmetrical, mixed lung pattern. Moderate bronchial and reticulonodular patterns affected the right lung lobes, partial collapse and an alveolar pattern were noted within the accessory lung lobe, and multiple larger well-defined nodules (some showing internal mineralisation) were present within the left lung lobes, with more normal appearing parenchyma surrounding them. There was moderate sternal and cranial mediastinal and marked tracheobronchial lymphadenomegaly. Histopathology on an incisional biopsy of the soft tissue mass revealed a large mixed inflammatory cell infiltrate including epitheloid macrophages, suggestive of mycobacteriosis; ZN staining revealed large numbers of acid-fast bacilli, which were identified by PCR as M microti.

Triple antibiotic therapy was instigated for 9 months (clarithromycin [11 mg/kg] 65 mg PO q12h; rifampicin [9 mg/kg] 50 mg PO q24h; marbofloxacin [1.8 mg/kg] 10 mg q24h). Within 2 months the stertor had resolved and the skin lesion had reduced in size; by the end of the 9 month course of antibiotic therapy all clinical signs had fully resolved. A CT scan showed improvement, but not resolution of the mediastinal and sternal lymphadenopathy and diffuse lung changes. The left lung nodules had mildly more extensive mineralisation than previously. It was decided to continue treatment due to the continued presence of pathology, and a timeline of this case is shown in Figure 4.

Discussion

The cases presented here are a cohort of cats with conclusive or strong evidence supporting a diagnosis of feline tuberculosis (culture, PCR and/or IGRA results). In contrast to previously published data on feline tuberculosis, the cases in this series are predominantly M microti infections, whereas national culture data shows that while M microti can be cultured from 19% of cases with histopathological changes indicative of mycobacteriosis, M bovis can usually be cultured from 15%. ² The reason for the lack of M bovis cases is unclear; it may be the result of our small sample size, the majority of which lived in regions of the UK where M microti is more prevalent, ² or it could indicate an underlying bias towards owners being more likely to treat cats with M microti-infection rather than M bovis, probably due to the higher zoonotic risk associated with the latter organism. ⁹

In line with previous studies, the majority of cats with tuberculosis in this study are males; ² none were found to be coinfected with feline immunodeficiency virus and feline leukaemia virus, and the median age of cats infected with M microti was 7 years (range 7 months to 13 years), compared with a previously documented median of 8 years. ²

The cases in this series demonstrated a range of clinical responses following diagnosis and treatment of disseminated feline tuberculosis, and in each case, repeated CT imaging contributed to decision making in ongoing clinical management within the context of contemporaneous investigations. It is recognised that the cases in this study show significant variability both in the use of CT and its timing in relation to treatment. This largely relates to the multi-centre nature of this study, as decision making varied depending on the preferences of the primary clinician.

A previous study found a sustained complete remission in only 40% of feline mycobacterial infections; ⁶ however, that study included many cases that were treated with sub-optimal drug regimens (eg, short courses of fluoroquinolone monotherapy),^6,10 as well as including M avium infections, which are known to be refractive to treatment due to complex inherent drug resistance patterns. ¹¹ Previously advocated treatment protocols for feline tuberculosis typically consisted of an initial and a continuation phase. ⁹ However, recent studies regarding multi-drug resistant M tuberculosis (MDR-TB) in humans have suggested that using at least three and ideally four antibiotics given in combination throughout treatment significantly reduces the development of antimicrobial drug resistance.^12–15

Recommended first-line anti-tuberculosis medications for humans consist of rifampicin, isoniazid, dihydrostreptomycin, ethambutol and pyrazinamide. ¹⁶ However, the use of these drugs does not readily translate into veterinary medicine; isoniazid has been associated with neurological side effects in small animals, ¹⁷ pyrazinamide is ineffective against M bovis infections, ¹⁸ which comprise approximately 15% of feline mycobacterial infections, ² and dihydrostreptomycin should be reserved for human use. ¹⁹ Therefore, small animal anti-tuberculosis therapy, when undertaken, should consist of a triple combination of rifampicin (for its potency and its ability to kill non-replicating [latent] tuberculous mycobacteria ²⁰ [recommended doses 10–15 mg/kg PO q24 h]), a fluoroquinolone (ideally pradofloxacin, as it has better efficacy against mycobacteria than older fluoroquinolones,^21,22 and a better safely profile in cats ²³ [pradofloxacin recommended doses 3–7 mg/kg PO q24h]) and a macrolide (such as clarithromycin [7–15 mg/kg PO q12h] or azithromycin [5–15 mg/kg PO q24h]) for a minimum of 3 months as standard.^9,24 It is recommended that treatment should be given for 2–3 months after apparent clinical resolution, which typically results in 4–6 months of treatment.^9,24

The efficacy of combination long-term treatment is supported by the cases in this series, as all were treated with either two or three antibiotics for at least 6 months; only one of the cats died from tuberculosis, and another was found to have latent tuberculosis after euthanasia for an unrelated disease. This gives a sustained complete remission rate of eight of nine cases (~90% remission), which is much higher than the 40% previously reported. ⁶ This is much more in line with our recent experiences, as following the introduction of sustained triple therapy the prognosis for feline tuberculosis appears to be closer to 70–80% success when treating cutaneous and/or pulmonary tuberculosis caused by M bovis or M microti (DGM and COH, unpublished data, 2016). Prolonged therapy is therefore recommended in all cases, and due care is required when advising clients on discontinuing treatment.

The majority of the cases in this series (cases 1, 4, 5, 7 and 8) demonstrated that where improvement in previously detected abnormalities can be identified on the basis of follow-up CT, tapering or cessation of treatment could be undertaken with greater confidence in the context of other clinical findings. However, for some of the cases (6 and 9) significant changes remained at follow-up CT, despite the cats being clinically well, and as a result triple antibiotic therapy was continued.

A previous study into the diagnostic and monitoring capacity of standard radiography in feline tuberculosis cases showed that with prolonged antibiotic therapy, detectable pathology is eliminated in the vast majority of cases. ⁴ By comparison, in this case series some of the abnormalities detectable by CT imaging remained present in the majority of cases, though not all cats underwent repeat imaging following complete cessation of treatment. It is likely that this discrepancy partly results from the greater sensitivity of CT in comparison with radiography for detection of milder changes, highlighting its value. However this must be considered when repeat CT imaging is used to decide whether antibiotic treatment can be discontinued; complete resolution of pulmonary pathology cannot be reliably anticipated, even with extended antibiotic therapy. This highlights the value of ongoing follow up imaging to document the lack of progression of changes, which can then be considered clinically incidental.

In some cats undergoing treatment for feline tuberculosis, periods of clinical and/or radiological remission can be followed by recurrence of clinical signs, sometimes on multiple occasions (as seen in cases 1 and 3). It is difficult to determine if this represents recrudescence of disease following subclinical infection (latency) in the intervening periods, or reinfection. For example, cats that are habitual hunters have repeated exposure to a population of infected prey (as is the case for the cat in case 1). The return of clinical disease may be associated with extremely subtle clinical signs (as in case 1). The associated CT abnormalities may be similarly subtle (as in Figure 1c), but when a radiologically normal appearance has been documented during the remission period, these subtle changes can be considered significant, allowing for prompt reintroduction of treatment. This case also demonstrates the importance of careful and dedicated patient observation on the part of the owners; monitoring sleeping respiratory rate is recommended in all cases of feline tuberculosis when undergoing treatment, even when there was initially no respiratory involvement.

When repeating diagnostic procedures, it is important to evaluate the potential benefit to the patient in relation to the costs involved. In the cases in this series, we feel that the major benefit is clear: namely, that the decision to either reduce/discontinue or restart treatment could be made with greater confidence. With reference to CT examination, a number of costs should be considered. The risk of repeated radiation exposure during scanning is one. We feel that in a population largely consisting of middle-aged cats the risk is minimal, though it should not be entirely discounted, particularly in cases where large numbers of repeated scans are performed. The effect of sedation or general anaesthesia should also be considered. Within a referral hospital the risks of these are low, ³⁰ but they may warrant consideration, particularly in clinically unstable patients with significant multisystem disease. Finally, the financial cost to the owner should also be considered. In several of the cases in this series, some of the associated costs and risks were reduced by use of a VetMouseTrap device, which allows for full body scanning in a non-sedated patient. Despite a slight reduction in sensitivity arising from a reduction in image resolution, this technique provides a very useful relatively low cost and non-invasive option. Notwithstanding the use of a VetMouseTrap device, in many referral centres the cost to the owner of a CT examination, either thorax in isolation or multiple body regions, does not significantly exceed that of full radiological examination. In addition, as CT becomes more widespread in non-specialist practice, its advantage as far as increased sensitivity over radiology warrants further consideration.

Conclusions

The cases described in this case series demonstrate the value of repeat CT imaging in the management of mycobacterial disease. When considered in combination with clinical findings, CT studies can aid in decision making regarding tapering of antibiotic protocols, or reintroduction of therapy in cases of recurrence or reinfection. This series also highlights that, in some cases, persistent abnormalities can be detected by CT, which may not necessarily indicate an active disease process, and care should be taken in the interpretation of these findings.