Equine paranasal sinus disease: A long-term study of 200 cases (1997–2009): Ancillary diagnostic findings and involvement of the various sinus compartments

Authors

  • P. M. DIXON,

    Corresponding author
    1. Division of Veterinary Clinical Studies, Royal (Dick) School of Veterinary Studies, The University of Edinburgh, Easter Bush Veterinary Centre, Midlothian, Scotland
    Search for more papers by this author
  • T. D. PARKIN,

    1. School of Veterinary Medicine, College of Medical, Veterinary and Life Sciences, University of Glasgow, Glasgow, UK.
    Search for more papers by this author
  • N. COLLINS,

    1. Division of Veterinary Clinical Studies, Royal (Dick) School of Veterinary Studies, The University of Edinburgh, Easter Bush Veterinary Centre, Midlothian, Scotland
    Search for more papers by this author
  • C. HAWKES,

    1. Division of Veterinary Clinical Studies, Royal (Dick) School of Veterinary Studies, The University of Edinburgh, Easter Bush Veterinary Centre, Midlothian, Scotland
    Search for more papers by this author
  • N. TOWNSEND,

    1. Division of Veterinary Clinical Studies, Royal (Dick) School of Veterinary Studies, The University of Edinburgh, Easter Bush Veterinary Centre, Midlothian, Scotland
    Search for more papers by this author
  • W. H. TREMAINE,

    1. Division of Veterinary Clinical Studies, Royal (Dick) School of Veterinary Studies, The University of Edinburgh, Easter Bush Veterinary Centre, Midlothian, Scotland
    Search for more papers by this author
  • G. FISHER,

    1. Division of Veterinary Clinical Studies, Royal (Dick) School of Veterinary Studies, The University of Edinburgh, Easter Bush Veterinary Centre, Midlothian, Scotland
    Search for more papers by this author
  • R. EALEY,

    1. Division of Veterinary Clinical Studies, Royal (Dick) School of Veterinary Studies, The University of Edinburgh, Easter Bush Veterinary Centre, Midlothian, Scotland
    Search for more papers by this author
  • S. Z. BARAKZAI

    1. Division of Veterinary Clinical Studies, Royal (Dick) School of Veterinary Studies, The University of Edinburgh, Easter Bush Veterinary Centre, Midlothian, Scotland
    Search for more papers by this author

email: P.M.Dixon@ed.ac.uk

Summary

Reasons for performing study: There is a lack of objective information on the value of ancillary diagnostic techniques used to investigate equine sinus disease, and also on which sinus compartments are commonly affected in this disorder.

Objectives: To record the ancillary diagnostic findings used to investigate equine sinus disease and to document which compartments are affected.

Materials and methods: The clinical case records of 200 consecutive cases of sinus disease, including subacute (<2 months' duration) primary (n = 52); chronic (>2 months' duration) primary (n = 37); dental (n = 40); traumatic (n = 13); sinus cyst (n = 26); sinus neoplasia (n = 10); dental related oromaxillary fistula (n = 8); mycotic sinusitis (n = 7) and intra-sinus progressive ethmoid haematoma (n = 7) were retrospectively examined.

Results: Nasal endoscopy showed exudate draining from the sino-nasal ostia in 88% of cases and a sino-nasal fistula was present in 15% of cases. Sinoscopy was performed in 79% of cases and was of great diagnostic value. More recently, 22% of cases had fenestration of the ventral conchal bulla performed to allow sinoscopy of the rostral sinus compartments. Radiography was performed in 97% of cases and showed intra-sinus fluid lines to be common (69% prevalence) in subacute primary sinusitis. Radiographic dental apical changes were not specific to dental sinusitis, e.g. 29% of chronic primary sinusitis cases had radiographic dental changes. Scintigraphy was performed in 20% of cases and was helpful in identifying dental apical changes when radiography was inconclusive. Overall, the caudal maxillary (78% involvement) and rostral maxillary (61%) sinuses were most commonly affected, with the ventral conchal sinus (VCS) (54% involvement) and conchofrontal sinuses (48%) less so. The VCS showed the greatest tendency to contain inspissated pus (present in 46% of all affected VCS).

Conclusions: Nasal endoscopy, sinoscopy and skull radiography are of great value in diagnosing the presence and causes of equine sinus disease.

Introduction

Sinus disease (commonly termed sinusitis) is a poorly understood, yet important equine disorder. The diagnostic techniques used to investigate equine sinus disease continue to evolve but there is a paucity of objective reports on the use or value of these techniques. There is also limited information on which compartments are affected by the different types of sinus disease and which tend to accumulate inspissated exudate. To gain further information on these areas, a retrospective study examined details of 200 cases of sinus disease. The treatments of these cases are presented in a companion paper (Dixon et al. 2012).

Materials and methods

A retrospective study was made of all cases of sinus disease referred to the Equine Hospital at the University of Edinburgh between January 1997 and April 2009. Diagnostic criteria for differentiating the different types of sinus disease were as previously described (Tremaine and Dixon 2001; Tremaine and Freeman 2007), except that in this study the data from cases of primary sinusitis were subdivided into subacute primary sinusitis (<2 months' duration) and chronic primary sinusitis (>2 months' duration) in order to examine for possible differences in clinical or ancillary diagnostic findings between these 2 groups. Also in this study, cases of sinusitis caused by dental disorders were divided into those caused by cheek tooth apical infection, i.e. classic dental sinusitis (Baker 1970) and those caused by dental-related oromaxillary fistula (OMF) formation (e.g. due to deep CT diastemata or defects in CT) that were termed dental-related OMF (Hawkes et al. 2008).

Identification of the individual compartments affected by sinus disease was made by radiography and confirmed in most cases by sinoscopy and/or direct inspection during sinusotomy (Dixon et al. 2012). A sinus compartment was defined as being affected by the presence of mucosal thickening or reddening, and/or by the presence of exudate (liquid or inspissated), food, blood or lesions (including progressive ethmoid haematoma [PEH], cystic, neoplastic or fungal plaque lesions) within its lumen.

Data analyses

Univariable and multivariable logistic regression models were developed to compare subacute with chronic primary sinusitis cases and to compare dental sinusitis and sinus cysts with combined primary sinusitis cases. Separate comparisons between the type of sinus disease and diagnostic findings were performed using Chi-square (or Fisher's exact) analysis where appropriate.

Results

Ancillary diagnostic techniques

Endoscopic, sinoscopic and radiographic findings are detailed in Table 1.

Table 1. Endoscopic, sinoscopic and radiographic findings in 200 cases of equine sinus disease
Type of sinus disease presentSubacute primary n = 52Chronic primary n = 37Dental n = 40Dental OMF n = 8Sinus cyst n = 26Traumatic n = 13Neoplasia n = 10Mycotic n = 7PEH n = 7
  1. Number in brackets = number of documented observations. Dental OMF = dental associated oromaxillary fistula; PEH = intra-sinus progressive ethmoid haematoma; s-n = sino-nasal; Rad. = radiographic; Rad. Bone changes = radiographic sinus wall bone changes.

Exudate at s-n ostium42 (50) 84%25 (35) 71%33 (38) 87%7 (8) 88%10 (20) 50%12 (12) 100%3 (4) 75%3 (5) 60%4 (6) 67%
Sino-nasal fistula8 (52) 15%7 (37) 19%6 (40) 15%1 (8) 13%2 (26) 8%1 (13) 8%1 (10) 10%1 (7) 14%2 (7) 29%
Sinoscopy performed42 (52) 81%33 (37) 89%29 (40) 73%8 (8) 100%23 (26) 88%6 (13) 46%3 (10) 30%7 (7) 100%7 (7) 100%
Open VCB sinoscopically16 (52) 31%8 (37) 22%11 (40) 28%2 (8) 25%2 (27) 7%0 (13) 0%1 (10) 10%1 (7) 14%0 (7) 0%
Radiography performed50 (52) 96%35 (37) 95%40 (40) 100%8 (8) 100%26 (26) 100%13 (13) 100%8 (9) 89%7 (7) 100%6 (7) 86%
Intra-sinus fluid lines34 (49) 69%16 (35) 46%16 (40) 40%3 (8) 38%8 (26) 31%11 (13) 85%1 (8) 13%2 (7) 29%2 (6) 33%
Intra-sinus radio-opacity26 (49) 53%25 (35) 71%22 (40) 55%6 (8) 75%23 (26) 88%4 (13) 31%5 (9) 56%4 (7) 57%6 (6) 100%
Rad. apical changes8 (49) 16%10 (35) 29%34 (40) 85%5 (8) 63%3 (26) 12%0 (13) 0%2 (9) 22%1 (7) 14%0 (6) 0%
Rad. bone changes3 (52) 6%2 (37) 5%1 (40) 3%1 (8) 13%2 (26) 8%7 (13) 54%0 (10) 0%1 (7) 14%1 (7) 14%

Nasal endoscopy

Sinus cysts extended through the sino-nasal ostium into the middle meatus in 2/26 (7.7%) horses with sinus cysts. Some horses had intermittent nasal discharge, and endoscopy of such cases when nasal discharge was absent often showed a small stream of exudate from the sino-nasal ostia being cleared in a caudo-ventral direction towards the nasopharynx by normal mucociliary action and gravity. Nasal endoscopy showed a sino-nasal fistula to be present in 15% of cases, with a similar distribution of this lesion between the different sinus disease categories (Table 1). These fistulas usually drained into the caudo-lateral aspect of the middle meatus and were often obscured by inspissated pus which, in some cases, contained pieces of lace-like, sequestered conchal bone. In some horses a sino-nasal fistula was only detected when this inspissated pus was removed by use of per nasum transendoscopic biopsy forceps and/or high-pressure transendoscopic lavage with up to 2 l of lukewarm water. In other cases, a fistula was only detected by performing endoscopy of the middle meatus, including examination of an area dorsal or dorso-lateral to the normal sino-nasal ostia drainage site, using a small (e.g. <8 mm) diameter endoscope. In a few cases, a sino-nasal fistula was not detected at nasal endoscopy but was observed following sinoscopy and lavage of inspissated pus from the VCS, when the endoscope tip was found to be lying in the nasal cavity.

Sinoscopy

Sinoscopy was performed in 79% of cases and was used primarily to identify which compartment had mucosal inflammation and/or contained lesions, food material or exudate (Tremaine 2007) (Table 1). Six cases had food material identified within their sinuses, 5 cases had sequestration of conchal bones (believed to be of ventral conchal bulla [VCB]) into the caudal maxillary sinus (CMS). One case had pieces of acrylic in its CMS (following dental repulsion prior to referral). Intra-sinus PEH lesions, blood clots in horses with head trauma and small sinus cysts were readily identified on sinoscopy, but sinoscopy of horses with large cysts that filled the conchofrontal sinus (CFS) resulted in the endoscope just entering the cyst lumen.

Fungal plaques directly attached to the sinus mucosa were readily detected on sinoscopy but fungal granulomas without an obvious overlying mycotic plaque were only identified histologically. Occasionally, fungal plaques or focal areas of black discoloration were found on the surface of masses of inspissated pus both within the sinus and at sino-nasal fistulas, and these were considered to be secondary fungal growths. Mycotic plaques that were considered secondary lesions were also found in conjunction with intra-sinus PEH lesions in 2 cases, and another PEH case developed a secondary mycotic infection after intra-lesional formalin treatment (Dixon et al. 2012).

Sinoscopy was seldom useful for making a definitive diagnosis of dental apical infection, as the generalised mucosal inflammation commonly present in affected sinuses obscured the possible detection of inflamed mucosa or granulomas present above some affected alveoli. Additionally, following ventral conchal bulla (VCB) fenestration it was usually not possible to fully examine the apical aspects of alveoli that lay in the rostral maxillary sinus (RMS) or VCS.

The VCB was opened in 41 (21%) more recent cases and the lumina of the VCS and RMS were subsequently examined (Table 1). This technique often confirmed the presence of inspissated pus in these compartments and sometimes allowed their transendoscopic treatment (Dixon et al. 2012). This technique was unsuccessful in 2 cases due to deformity of the VCB. Occasionally, excessive haemorrhage occurred when opening a thickened or inflamed VCB that precluded immediate examination of the VCS or RMS, which were examined the following day.

Radiography

Sinus radiography was performed in 97% of cases. Fluid lines, indicative of intra-sinus liquid pus (except where affected sinuses contained only blood - Table 1), were present in 48% of these cases and were significantly less common with chronic primary sinusitis (46%) than subacute primary sinusitis (69% prevalence) (OR 0.4; CI 0.17–0.96 P = 0.04). Compared to combined primary sinusitis cases, fluid lines were significantly less commonly associated with sinus cysts (31%) (OR 0.4; CI 0.16–0.96, P = 0.04). A diffuse radio-opacity of the affected sinuses was present in 61% of cases and cases with sinus cysts were 4.3 times more likely to have a generalised sinus opacity (95% CI 1.4–13.4, P = 0.01) than were cases of primary sinusitis. Definitive radiographic changes of the external sinus wall (bone sclerosis, lysis or distortion of the normal contour) were recorded in just 9% of cases, including in 54% of traumatic sinusitis cases. Radiographic changes of the apices of the intra-sinus cheek teeth were recorded in 31% of all cases, including in horses without dental-related sinusitis, including in 16% of subacute primary and 29% of chronic primary sinusitis cases. As expected, higher levels of such changes were present in cases of dental sinusitis (85%) and dental-related OMF (63%).

Scintigraphy

Scintigraphy was performed in 39 (19%) cases, usually to differentiate between primary and dental sinusitis when clinical and radiographic examinations were inconclusive. Most (72%) cases of primary sinusitis examined by scintigraphy showed increased radionucleotide uptake (IRU) localised to bone other than perialveolar bone (i.e. to maxillary, lacrimal, malar, frontal, nasal and conchal bones) (Table 2). Additionally, focal areas of marked IRU of nonalveolar bone were present in 19% of primary sinusitis cases that sometimes overlaid the dental apices on lateral projections, but on other projections these areas of IRU were clearly shown not to lie adjacent to dental apices. Mild IRU of alveolar bone was also detected in 28% of primary sinusitis cases, and this IRU usually involved multiple alveoli compared to cases with confirmed apical infections.

Table 2. Scintigraphy findings in 39 sinus disease cases that underwent scintigraphy
Type of sinus diseaseSubacute primary n = 52Chronic primary n = 37Dental n = 40Dental OMF n = 8Sinus cyst n = 26Traumatic n = 13Neoplasia n = 10Mycotic n = 7PEH n = 7
  1. Dental OMF = dental-related oromaxillary fistula; PEH = intra-sinus progressive ethmoid haematoma; Alveolar IRU = Marked alveolar bone increased radionucleotide uptake; Bone IRU = nonalveolar bone IRU; Alveolar and bone IRU = combined IRU of alveolar and nonalveolar bone.

Scintigraphy performed11 (52) 21%7 (37) 19%9 (40) 23%1 (8) 13%7 (26) 27%2 (13) 15%0 (10) 0%1 (7) 14%1 (7) 14%
Alveolar IRU0 (11) 0%0 (7) 0%5 (9) 56%0 (1) 0%0 (7) 0%0 (2) 0%0 (10) 0%0 (1) 0%0 (1) 0%
Bone IRU9 (11) 82%4 (7) 57%0 (9) 0%0 (1) 0%7 (7) 100%2 (2) 100%0 (10) 0%1 (1) 100%1 (1) 100%
Alveolar and bone IRU2 (11) 18%3 (7) 43%4 (9) 44%1 (1) 100%0 (7) 0%0 (2) 0%0 (10) 0%0 (1) 0%0 (1) 0%

Microbiology

See supplementary information (Appendix S1)

Histology

Histology of sinus lesions was performed using previously described techniques and diagnostic criteria (Dixon and Head 1999; Tremaine et al. 1999).

Sinus compartments affected

Overall, the CMS was the most commonly affected compartment (affected in 78% of all cases) (Table 3). Inflammation of the smaller SPS and ES sinuses was uncommonly identified (6% recorded as being affected) in this study. Sinus cysts tended to involve the more caudal sinus compartments and the multivariable model showed that the RMS was 2.7 times more likely to be affected in cases of primary sinusitis than in cases of sinus cysts (OR 2.7, 95% CI 1.01–7.36, P = 0.05).

Table 3. Proportions of individual sinus compartments affected in each disorder and proportions containing inspissated (inspiss) pus
Type of sinus diseaseSubacute primary n = 52Chronic primary n = 37Dental n = 40Dental OMF n = 8Sinus cyst n = 26Traumatic n = 13Neoplasia n = 10Mycotic n = 7PEH n = 7All 200 cases
  1. Dental OMF = dental related oromaxillary fistula; PEH = intra-sinus progressive ethmoid haematoma; CFS = conchofrontal sinus; CMS = caudal maxillary sinus; RMS = rostral maxillary sinus; VCS = ventral conchal sinus; SPS = sphenopalatine; ES = ethmoidal sinus.

CFS31 (52) 60%22 (37) 59%12 (40) 30%3 (8) 38%11 (26) 42%5 (13) 38%5 (9) 56%4 (7) 57%2 (7) 29%95 (48%)
CMS42 (52) 81%31 (37) 84%29 (40) 73%7 (8) 88%21 (26) 81%9 (13) 69%6 (9) 67%6 (7) 86%5 (7) 71%156 (78%)
RMS36 (52) 69%26 (37) 70%31 (40) 78%3 (8) 38%10 (26) 38%6 (13) 46%5 (9) 56%2 (7) 29%2 (7) 29%121 (61%)
VCS31 (52) 60%23 (37) 62%30 (40) 75%4 (8) 50%12 (26) 46%1 (13) 8%2 (9) 22%2 (7) 29%3 (7) 43%108 (54%)
SPS/ES4 (52) 8%2 (37) 5%2 (40) 5%2 (8) 25%1 (26) 4%0 (13) 0%0 (10) 0%1 (7) 14%0 (7) 0%13 (7%)
CFS inspiss pus2 (52) 4%6 (37) 16%1 (40) 3%4 (8) 50%0 (26) 0%0 (13) 0%0 (10) 0%0 (7) 0%0 (7) 0%13 (7%)
CMS inspiss pus11 (52) 21%12 (37) 32%8 (40) 20%4 (8) 50%1 (26) 4%0 (13) 0%0 (10) 0%3 (7) 43%0 (7) 0%39 (20%)
RMS inspiss pus15 (52) 29%12 (37) 32%6 (40) 15%0 (8) 0%0 (26) 0%0 (13) 0%0 (10) 0%0 (7) 0%0 (7) 0%33 (17%)
VCS Inspiss pus19 (52) 37%16 (37) 43%8 (40) 20%2 (8) 25%2 (26) 8%1 (13) 8%0 (10) 0%2 (7) 29%0 (7) 0%50 (25%)
SPS/ES inspiss pus2 (52) 4%2 (37) 5%2 (40) 5%0 (8) 0%0 (26) 0%0 (13) 0%0 (10) 0%0 (7) 0%0 (7) 0%6 (3%)

The proportion of sinuses that contained inspissated pus also differed greatly between compartments, e.g. inspissated material was present in 46% of affected VCS compartments, in contrast to just 14% of affected CFS compartments. Intra-sinus inspissated pus was found with many types of sinus disease (including in the VCS of 20% of dental sinusitis cases), but was an uncommon finding with sinus cysts, and the multivariable model found that the CMS was significantly less likely to contain inspissated pus in horses with sinus cysts than those with primary sinusitis (OR 0.10; CI 0.01–0.97; P = 0.047).

Discussion

Nasal endoscopy

The detection of exudate draining from the sino-nasal ostium by nasal endoscopy was a simple and accurate method of confirming the presence of sinusitis. This finding was present in 78% of all cases where examination of this area was possible. The presence of a nonsurgically created sino-nasal fistula has been poorly described in equine sinus disease, but was found in 15% of current cases and some images of these lesions have recently been published (O'Leary and Dixon 2011). These fistulas are likely to be a sequel to chronic infection and distension of the VCS, RMS or, less often, the DCS or CMS, resulting in necrosis of some aspects of their walls. In support of this hypothesis, sequestered pieces of nasal concha were present in the nasal cavities, adjacent to some fistulas, and were also found at sinusotomy in the CMS of some horses with chronic primary sinusitis (Dixon et al. 2012). The detection of sino-nasal fistulas at nasal endoscopy usually allows sinoscopy to be performed via the nasal cavity and allowed treatment of some cases via this route (Dixon et al. 2012). Sinus (sino-nasal) cysts were found extending through the sino-nasal ostium into the middle meatus in 25% of affected horses by Woodford and Lane (2006) but in only 5% of current cases.

Sinoscopy

Sinoscopy was found to be a minimally invasive technique that enabled the visualisation of inspissated pus in some primary sinusitis cases, explaining their chronicity and non-response to sinus lavage. It also helped identification of intra-sinus PEH and mycotic lesions and the detection of food in the sinus. In agreement with Perkins et al. (2009b), sinoscopy was rarely useful in diagnosing dental sinusitis, although Perkins and Schumacher (2010) have published sinoscopic images of dental sinusitis-related granulomas.

Endoscopic examination of the RMS and VCS was usually possible following perforation of the VCB, without having to trephine directly into the RMS - which is technically difficult in young horses and risks damaging the alveoli (Barakzai et al. 2008; Perkins et al. 2009a). Ventral conchal bulla fenestration also allowed sinoscopically guided treatment of RMS or VCS compartments and post operative lavage of all sinus compartments via a single (frontal) portal. The use of a 3 mm diameter endoscope to perform intra-sinus endoscopy via the naso-maxillary ostium (M Nowak 2009 personal communications) or the use of sinoplasty of the sino-maxillary ostium as described in vitro by Bell et al. (2009) are less invasive techniques to examine and treat some sinusitis cases.

Radiography

Sinus radiography was the most commonly performed ancillary diagnostic technique in this study and showed the presence of fluid lines in 48% of all cases, whereas Feige et al. (2000) found this feature in 85% of horses with sinusitis. Fluid lines are indicative of liquid pus (or of blood in traumatic sinusitis and intra-sinus PEH cases) and were significantly more common in cases of subacute (69% prevalence) than chronic primary sinusitis (46%), presumably because, in chronic cases, the exudate had become inspissated. Compared to (combined) primary sinusitis cases (60% prevalence), fluid lines were less commonly present in cases of dental sinusitis (40%) or sinus cysts (38%). Gibbs and Lane (1987) found a very similar prevalence of fluid lines present in primary sinusitis (59%) and dental sinusitis cases (43%).

The diffuse radio-opacity of the sinuses present in 63% of all cases likely reflects bone and mucosal thickening, as well as the presence of intra-sinus inspissated pus. In support of this hypothesis, this feature was slightly more common in chronic (71% prevalence) than in subacute (53%) primary sinusitis and was also very commonly identified in cases of sinus cysts (88%) and dental sinusitis (75%).

Interpretation of cheek teeth radiographic apical changes has a low sensitivity (Gibbs and Lane, 1987; Tremaine and Dixon, 2001; Weller et al. 2001; Townsend et al. 2011). It can be particularly difficult to detect apical changes in cases with concurrent sinusitis, owing to the superimposition of radio-opaque sinus contents or thickened sinus walls on the dental apices (Gibbs and Lane 1987; Weller et al. 2001). Some radiographic apical changes of the intra-sinus cheek teeth were recorded in 32% of all cases, including in 29% of chronic primary sinusitis cases, as well as in 85% of dental sinusitis and 50% of dental-related OMF cases (with definitive radiographic dental changes sometimes identified in the latter 2 categories).

Scintigraphy

Scintigraphy was usually performed to detect the presence of apical infection, with affected teeth having a marked, local increase in IRU at the apex of the affected tooth (Weller et al. 2001; Barakzai et al. 2006). Focal areas of marked IRU of bone were also present in 19% of primary sinusitis cases (in addition to 72% of cases that had diffuse IRU of bone), which had to be distinguished from alveolar IRU by a combination of scintigraphic projections. Mild IRU of alveolar bone was detected in 28% of primary sinusitis cases but usually overlaid multiple alveoli and was of lesser intensity than that observed with confirmed apical infections. Long-term follow-up of the latter cases (Dixon et al. 2012) showed that they resolved without dental extraction, providing clinical evidence that apical infection was not the cause of their sinusitis.

Microbiology

See supplementary information (Appendix S1)

Sinus compartments affected

This survey showed the maxillary sinuses to be most commonly affected with sinus disease, with the CMS and RMS affected in 78% and 61%, respectively, of all 200 cases. The VCS (54%) and CFS (48%) were less commonly affected (Table 3). SPS and ES involvement was uncommonly diagnosed; however, inflammation of these 2 compartments is not easily distinguished using radiography, and a specific effort needs to be made to examine the SPS endoscopically. Ethmoidal sinusitis was not identified in this study unless gross changes were present on the external surfaces of the second ethmoturbinate or exudate was observed draining from its caudal aspect at nasal endoscopy. It appears likely that many cases of sinusitis involving the CMS have undiagnosed inflammation of the interconnected SPS and ES. The increased use of CT will more accurately quantify the involvement of the various compartments in equine sinus disease.

In general, intercommunicating sinus compartments were affected with similar levels of sinus disease. However, in some cases, involvement of the the CMS (78% involvement of all cases) but not of the more dorsally positioned (and hence better draining) CFS (48% involvement) was recognised. Likewise, the RMS (61%) was slightly more commonly affected than the VCS (54%). Although the latter 2 compartments should communicate freely dorsal to the infraorbital canal, the communication between them can be quite narrow (1–8 mm), particularly in young horses (Probst et al. 2005). Consequently, in horses with sinusitis, when the inflamed sinus mucosa (which lies above and below this ostium) can become <15 mm thick (Tremaine et al. 1999), the VCS can become sealed off from the RMS by mucosal inflammation.

In 32 cases of primary sinusitis, Quinn et al. (2005) recorded involvement of the RMS and VCS (single value given for both) in 71% and of the CFS and CMS (single value also given) in 56% of cases, which differs from the current findings in cases of (combined) primary sinusitis, where the RMS (70% affected) and VCS (61%) were not more commonly affected than the CMS (76%). In the current study, sinus cysts tended to involve the more caudal sinus compartments, especially the CMS, with the RMS less commonly affected, as previously recorded (Tremaine and Dixon 2001; Woodford and Lane, 2006).

The presence of intra-sinus inspissated pus has largely been associated with primary sinusitis (Schumacher et al. 1987; Freeman et al. 1990). However, in this study, intra-sinus inspissated pus was found to be present with many other types of sinus disease (Table 3). This study also showed a very strong tendency for an affected VCS to contain inspissated exudate as previously recorded (Schumacher et al., 1987; Freeman et al. 1990; Quinn et al. 2005; Perkins et al. 2009b). In the current study, 46% of VCS affected with any type of sinus disease contained inspissated pus, in contrast to just 14% of affected CFS or 27% of affected RMS compartments. The VCS is the most dependent sinus, containing many diverticula (Perkins 2002) and has only indirect communication with the nasal cavity via the RMS. These factors may restrict VCS drainage and predispose any exudate it contains to inspissation.

Once inspissated pus is present in a sinus compartment, it is extremely unlikely that the sinusitis will resolve with medical treatment. Additionally, sinus lavage via CMS or CFS portals is an ineffective treatment for VCS or RMS infections because, with chronic sinusitis, the maxillary septum usually remains intact, preventing communication between the rostral (VCS and RMS) and caudal groups of sinuses (CFS and CMS) (Dixon et al. 2012). However, as noted earlier, the VCB is disrupted in a small proportion of chronic sinusitis cases.

Conclusions

In conclusion, this study found nasal endoscopy, frontal sinoscopy and radiography to be of great value in diagnosing the presence of and identifying the causes of equine sinusitis. Nasal endoscopy showed exudate at the sino-nasal ostium to be a feature of most cases and showed 15% to have a pre-existing sino-nasal fistula. All sinus compartments, especially the CMS, can be involved in sinus disease, and the VCS shows the greatest tendency to retain inspissated exudate.

Authors' declaration of interests

No conflicts of interest have been declared.

Sources of funding

This study was funded by the Division of Veterinary Clinical Studies, The University of Edinburgh.

Acknowledgements

During this study, the clinical residencies of Safia Barakzai, Niamh Collins and Claire Hawkes were generously funded by the Horserace Betting Levy Board.

Author contributions The initiation, conception planning were by P.M.D. All authors helped with execution. P.M.D. and S.B. performed the writing. The statistics were by T.D.P.

Ancillary