Introduction

Odontogenic sinusitis (OS) is a common, often overlooked disease [1]. Computed tomography (CT) plays a key role in revealing OS [2] and remains the method of choice as an imaging modality in chronic rhinosinusitis patients [3]. Cone beam computed tomography (CBCT), featuring a lower radiation dose, but a poorer view of soft tissues, can serve as an alternative [4]. In addition, CT and CBCT provide valuable data on the adjacent dentition and maxillary bone.

Radiological reports from rhinosinusitis patients frequently exclude dental pathology [5], and otolaryngologists and rhinologists have stated that radiologists rarely comment on dental pathology in CT scans [6]. Insufficient radiological reports may mislead the diagnosis and delay appropriate treatment and recovery. Untreated OS causes burdens to healthcare systems and can lead to persistent symptoms or even life-threatening intra-cranial complications.

Various dental conditions and procedures can lead to OS [7]. Apical periodontitis [8] and alveolar maxillary bone loss [9] associate with sinus mucosal thickening. Moreover, sinonasal complications often result from dental treatment, such as tooth extractions, endodontics, and implant surgery including alveolar ridge augmentation procedures [10]. Occasionally, maxillary sinus disorders are linked to peri-implantitis, as well as ectopic teeth and related dentigerous cysts [11]. In the case of a rhinosinusitis patient, most dental pathologies must be considered as possible causative agents.

Effective treatment protocols and several successful case series on managing OS patients have been published [10, 12]. Utilising these, however, calls for determining the correct diagnosis and the specific dental cause of disease.

Here, we aimed to investigate (i) how often and (ii) with which precision dental pathology is evaluated in the original CT/CBCT reports of rhinosinusitis patients and (iii) whether some findings are overlooked. Additionally, we studied how often an insufficient area of interest (AOI) or artefacts hinder the accurate interpretation of radiological findings concerning teeth. Our primary aim was to improve OS radiological diagnostics and identify areas needing more attention.

Materials and methods

Data were retrospectively obtained for patients who visited a tertiary hospital, namely, Helsinki University Hospital’s (HUH) Department of Otorhinolaryngology, in 2013 because of acute or chronic rhinosinusitis.

The records of all 2366 patients with International Statistical Classification of Diseases and Related Health Problems tenth revision (ICD-10) code J32 (chronic sinusitis) or J01 (acute sinusitis) including subcodes for maxillary, frontal, sphenoidal, ethmoidal, pansinusitis, other, and unspecified sinusitis were first scrutinised. Patients with clearly isolated frontal sinusitis and sphenoidal sinusitis were excluded because of their unlikely association with odontogenic factors. In addition, patients with previous visits to HUH before 2013 because of rhinosinusitis or with a known history or clear radiological proof of sinus surgery were excluded. From this patient cohort we included patients aged >18 years with CT/CBCT examination and its original report electronically available. Fig 1 illustrates the patient selection. CT/CBCT images and their original reports were used as the study subjects.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Flowchart of patient selection.

https://doi.org/10.1371/journal.pone.0299489.g001

Radiological examination was performed either at the referring clinic or in our tertiary healthcare setting. Examinations were originally interpreted by a specialist in radiology or a resident. Following interpretation by a resident, a specialist had confirmed the report.

Data collection

All comments on the maxillary teeth and dental pathology were obtained from the original reports. Findings were categorised per tooth as precisely as possible. We counted reports in which OS was suspected or excluded.

An oral and maxillofacial (OMF) radiologist with more than 20 years of clinical experience and blinded to the original reports re-evaluated all CT/CBCT scans. The images were analysed on a Barco E-3620 3MP Medical Flat Grayscale Display (Barco, Kortrijk, Belgium) with a diagonal dimension of 20.8 inches and a resolution of 2.048 × 1.536 pixels with a pixel size of 0.207 mm.

Table 1 lists the dental findings the OMF radiologist sought.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Dental findings assessed by an oral and maxillofacial radiologist.

https://doi.org/10.1371/journal.pone.0299489.t001

We compared the most typical OS dental findings from the original CT/CBCT reports and the OMF radiologist’s re-evaluation and determined the level of agreement between these findings (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Dental findings from computed and cone beam computed tomographies of rhinosinusitis patients (n = 300).

Comparison of original interpretations and oral and maxillofacial (OMF) radiologist’s re-evaluation.

https://doi.org/10.1371/journal.pone.0299489.t002

Results

In total, 296 CT and 4 CBCT images with the original interpretations were available for re-evaluation. The rhinosinusitis patient cohort consisted of 196 female (65.3%) and 104 male (34.7%) patients with a mean age of 45.2 ± 16.7 years.

Dental findings in the OMF radiologist’s re-evaluation

According to the OMF radiologist’s re-evaluation, among 225 examinations originally lacking any reported dental pathology, 22 had apical periodontitis, 16 had severe alveolar bone loss, 4 had pulp perforating caries, 16 had underfilled root canals, 11 had unfilled root canals, 2 had overfilled root canals, and 1 had a fractured root.

Table 2 summarises the dental findings commonly associated with OS and compares these with the OMF radiologist’s re-evaluation. Figs 2 and 3 provide examples of dental findings not mentioned in the original reports.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2.

Overlooked dental findings in maxillary molars (marked with arrows) from computed tomography scans: (a) coronal view of an apical periodontitis; (b) sagittal view of a root canal overfilling in a palatal root generally regarded as harmless, mesiocentral root canal with a suboptimal root canal filling with apical periodontitis (not shown) and a suspected reactive retention cyst; and (c) axial view of an untreated mesiocentral root canal in a mesiobuccal root.

https://doi.org/10.1371/journal.pone.0299489.g002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3.

Overlooked dental findings in maxillary molars (marked with arrows) from computed tomography scans: (a) sagittal view of a pulp perforating caries and (b) sagittal view of severe marginal alveolar bone loss.

https://doi.org/10.1371/journal.pone.0299489.g003

In the canine–molar region, dental pathology was identified in 102 patients (34.0%) with apical periodontitis emerging as the most common dental pathology. At least 1 root canal treated maxillary tooth was found in 88 patients (29.3%), among whom 27 (9.0%) had at least 1 untreated root canal. Apical periodontitis, marginal alveolar bone loss, and root canal treatment deficiencies typically related to the molars.

Because of the technical limitations of the CT/CBCT scans, diagnosing caries was not feasible on both sides for 219 patients. Overall, 18% of the examinations covered maxillary tooth crowns and marginal alveolar bone. These structures were either partly or not at all visible in the remainder of the examinations. Almost one-third (31.0%) had an artefact that hindered evaluation of the dental pathology. Moreover, 145 examinations (48.3%) had no artefact, but their AOI was inadequate to diagnose the crown and marginal alveolar bone.

Finally, 2 patients had a root remnant perforating the oral cavity and 1 patient had an infected root remnant inside the maxillary bone.

Discussion

In our cohort of rhinosinusitis patients’ CT/CBCT images and their original radiological reports, information about dentition was infrequently and briefly reported, largely restricted to apical periodontitis. Furthermore, among 225 patients whose original radiological reports did not mention dental pathology, 22 patients had apical periodontitis, 16 had severe alveolar bone loss, and 4 had pulp perforating caries according to the OMF radiologist. Agreement on sinus perforating apical periodontitis and non-perforating apical periodontitis were only moderate. These findings show that important radiological dental findings possibly causing sinonasal symptoms can be easily overlooked.

Apical periodontitis was the most frequent dental pathology, present in approximately one-fifth of the patients. The majority of the lesions did not perforate the sinus floor radiologically, rendering assessment of their possible role in sinus symptoms more challenging. Infection originating from apical periodontitis can actually advance into the sinus through the maxillary bone via the lymph and blood vessels without any bone destruction occurring [13]. Therefore, a radiologically intact cortex can be found between the apical lesion and maxillary sinus in cases of OS, and the affected tooth should be evaluated clinically by a dentist. CT provides valuable information also for the dental procedure.

In almost half (48.3%) of the images, AOI did not cover the tooth crown and marginal alveolar bone for diagnostics. According to the OMF radiologist, severe alveolar bone loss was, nevertheless, visible in 11.0% of the images, and thus originally significantly underreported. Maxillary sinus mucosal thickening associates with moderate and severe alveolar bone loss [9], and periodontitis associates with chronic rhinosinusitis [14]. Therefore, severe alveolar bone loss should also be seriously considered and treated as a possible condition underlying a patient’s sinus symptoms.

Artefacts hindered the interpretation of dental findings in 31.0% of cases. One artefact in CT/CBCT is a distortion or error in an image unrelated to the studied subject, which significantly affects the evaluation of pathologies such as caries [15]. Notably, pulp-extending caries can cause sinus mucosal changes before radiologically visible apical periodontitis is observed. Hence, as artefact may limit the utility of CT/CBCT for dental pathology, also 2-dimensional images have a role for OS diagnostics when pulpal necrosis is suspected.

As expected, the pathological dental findings in this study most often related to the maxillary molars. Compared to the premolars, the maxillary molars often result in root canal treatment failures [16]. Additionally, the molar roots accompany the shortest distance to the maxillary sinus floor. The molars are also most likely to have apical bone defects [17] and develop periodontitis. Teeth with an inadequate root canal filling are more likely to present with apical periodontitis [18]. In our study, at least one root canal treated maxillary tooth was found in 29.3% of patients, among whom almost one-third had at least 1 untreated root canal. Although root canal overfilling does not directly cause rhinosinusitis, resistance to infections may be weakened because of changes in the sinus mucosa.

The main characteristic of the reports studied was the sparse comment on the maxillary teeth. According to the OMF radiologist, dental pathology was observed in over one-third of the patients in the maxillary canine–molar region. In original reports, OS was apparent according to a radiologist in 8 of 300 (2.7%), excluded in 3 of 300 cases, and only 1 report from 300 recommended consulting a dentist. Considering the relevance of the possible dental finding was thus primarily left to the ear, nose, and throat specialist in question.

A standardised and structured format improves the quality of radiological reporting [19]. Furthermore, attention should be paid to using accurate terminology. A more thorough report of dental findings could simplify the process of determining the treatment path. We have previously suggested that dental findings could be primarily categorised as follows: (1) radiological findings referring to OS, (2) potential radiological findings referring to OS, and (3) no radiological findings referring to OS. Consulting a dentist should always be included in the first two categories [20]. In addition, the consulting dentist should have access to all relevant radiological examinations.

The strength of this study lies in the usability of these results to improve OS diagnostics and clinical practice guidelines [21]. Some limitations to this study should also be acknowledged. First, re-evaluation was performed by a single OMF radiologist in a single centre. Second, because of the retrospective nature of our study and our inability to verify the diagnosis, we could not categorise pathological dental findings as an obvious or potential cause of OS nor as directly causing sinonasal symptoms.

Conclusions

In relation to dental findings, computed tomography and cone beam computed tomography reports for rhinosinusitis patients remained limited and cursory, whereby alveolar bone loss was particularly overlooked. In our study, both apical periodontitis and severe alveolar bone loss were originally significantly underdiagnosed. The radiologist’s contribution to odontogenic sinusitis diagnostics or reporting alleged pathological dental findings as a cause of sinonasal symptoms–or alternatively as only co-existing findings—is essential. Therefore, comments on dentition visible in the optimised area of interest should always be included in a sinus computed tomography report.

Acknowledgments

Statistical analysis was performed by statistician Tero Vahlberg.