Tonsilloliths are oropharyngeal concretions stemming from a reactive foreign nidus such as bacteria and organic debris within a palatine tonsillar crypt. Based on empirical knowledge, tonsilloliths are reported to be relatively commonly encountered in daily clinical practice, but patients rarely have complaints related to them. It has been reported that the detection rate of tonsilloliths was under 25% [1–4], although experience suggests otherwise. Furthermore, recent studies suggest that tonsilloliths are clinically related to halitosis and tonsillar abscess [5–7]. Therefore, the detection rate, number, size, location, calcification level, and other characteristics of tonsilloliths should be precisely understood. However, CT was not used and/or the sample size was relatively small in the previous reports [1–4].

In the present study, the detection rate, number, size, location, and calcification level of tonsilloliths were analyzed by CT images. In addition, the possible causes of the low detection rates on panoramic radiographs were elucidated based on comparing the CT images and panoramic radiographs to elucidate the limitations of visualizing the area around the palatine tonsils on panoramic radiographs.

This study was based on 482 pairs of panoramic radiographs and CT scans that were obtained at the Division of Oral and Maxillofacial Radiology in Kyushu Dental University Hospital between 2010 and 2011 from patients with oral and maxillofacial region diseases (201 males, 281 females; mean age, 52.9 years; range, 5 to 94 years). Approval of the present study was obtained from the institutional review board of Kyushu Dental University (No. 12-19).

Panoramic radiographs were acquired using a panoramic AUTO-1000 EX system (Asahi Roentgen Ind. Co., Ltd., Kyoto, Japan). Images were taken in the incisive occlusion position holding the head by head supports with the FH plane parallel to the ground. CT was performed with an Activion 16 (Toshiba Co. Ltd., Tokyo, Japan). CT images were taken in the axial plane with 3-mm-thick contiguous sections from under the level of the orbit to under the hyoid bone to evaluate approximately the same regions viewed on panoramic radiographs. Images were obtained with standard algorithms and soft-tissue windows.

The presence of tonsilloliths was evaluated both on panoramic radiographs and CT images independently. Using the differential diagnosis of tonsilloliths on panoramic radiographs reported by Ram et al, a radiopaque nodular mass or masses piled up on the mandibular ramus and soft palate were defined as tonsilloliths [8]. The 482 pairs of images were assessed by a single, experienced dental radiologist (M. O.). The radiologist then assessed the CT and the panoramic radiographs of the same patients separately, without knowing the relationship between the two.

At the same time, the numbers, shapes, size, locations, and calcification levels of the tonsilloliths were also evaluated on both panoramic radiographs and CT scan. The masses or high density structures within the palatine tonsil were counted on the images. On CT scan, the tonsilloliths were diagnosed based on structures with over 100 Hounsfield Units (HU). The long axes of the radiopaque masses or high-density structures were measured. If the subject had multiple tonsilloliths, the size, shapes, location, and calcification levels of the largest calculus were analyzed. The locations of the tonsilloliths were divided into 3 groups (under the soft palate; coincident with the soft palate; over the soft palate) on the right or left, respectively. At the same time, the locations were divided into 2 groups, the medial and lateral halves of the lateral palatine tonsil determined based on the midline of the lateral palatine tonsil. The calcification levels of the tonsilloliths were evaluated as the maximum HU value in the density changing area.

Next, in cases of disagreement between the CT images and panoramic radiographic findings, the causes for disagreement were analyzed based on the CT findings as the gold standard. The difference in the numbers, shapes, size, locations, and calcification levels of the tonsilloliths were examined using statistical analyses of the cases in which there were and was not agreement.

All statistical analyses were performed using SPSS version 11 statistical software (SPSS, Chicago, Illinois, USA). Categorical variables were compared by the χ² test. The relationships between categorical variables were assessed using Pearson’s correlation coefficient or Spearman’s rank correlation. Results were considered significant if p < 0.05.

One of the important results of the present study was that relatively many subjects with no symptoms in the oropharyngeal area had one or more tonsilloliths (46.1%). Some investigators have also previously reported that the detection rate of tonsilloliths was very low [1, 2]. However, the data from these reports did not coincide with our clinical empirical knowledge based on CT images. This may be explained by the fact that CT images were not used in most of the previous reports on tonsilloliths [1, 2]. To the best of our knowledge, the detection rate of tonsilloliths was relatively high (about 15% or 25%) in only two reports using CT images [3, 4]. The sample size was relatively low (100 or 150 subjects), so that precise analysis could probably not have been done. In addition, the tonsillolith detection rate in the present study was significantly higher in subjects over 40 years old. However, the detection rates within the under 40-year-old subjects and the over 40-year-old subjects were relatively stable. In the previous reports, the conclusions on the relationship between the tonsillolith detection rate and age were contradictory. One concluded that the tonsillolith detection rate was related to age [9], while the other did not [10]. The present precise results can explain the contradictory opinions suggested by the previous reports; a very significant difference was observed in the relationship between the detection rate and age in the over and under 40-year-old groups. On the other hand, the correlation coefficient between the two may be relatively low, because the detection rates within the under 40-year-old subjects and within the over 40-year-old subjects were relatively stable.

Another of the important results is that the tonsilloliths detected on panoramic radiographs represent only a small part of the tonsillar calculi that were actually present. The tonsillolith detection rate on panoramic radiographs was only 7.3%, about one-sixth of that on CT images. The present results indicate that the main causes for the discrepancy were the calcification levels and the sizes of the tonsilloliths. In fact, calculus sizes were <5 mm in 179 of the 187 cases of tonsilloliths undetected on panoramic radiographs. In addition, almost all cases with CT number < 300 HU were not detected on panoramic radiographs. At the same time, the tonsillolith detection rate on panoramic radiographs was also significantly related to the number of tonsilloliths.

However, the causes for the discrepancy between panoramic radiographs and CT could not be elucidated in 3 cases. In 3 cases, the tonsilloliths were at a relatively low level that coincided with the mandibular roots and the heads of the submandibular glands. The tonsilloliths may have been obscured by the teeth and trabecular bones.

The location of tonsilloliths was not significantly related to the tonsillolith detection rate on panoramic radiographs. However, enostosis in the mandibular ramus was misdiagnosed as tonsilloliths on panoramic radiographs in two cases of disagreement between CT and panoramic radiographs. On panoramic radiographs, the palatine tonsil overlaps the mandibular ramus. It might be very difficult to differentiate between tonsilloliths and calcification diseases involving the mandible, such as enostosis. Therefore, based on the present results, panoramic radiographs cannot be considered to be a useful screening tool for tonsilloliths in the general dental population. In particular, CT examinations might be used for patients with clinical findings such as halitosis and a swollen palatine tonsil of unknown origin. At the same time, the characteristics of tonsilloliths on panoramic radiographs were accurately observed. The characteristic findings on panoramic radiographs of tonsilloliths were radiopaque mass(es) of various size(s) (mainly from 5 mm to 10 mm) near the soft palate, overlapping the mandibular ramus. In addition, no tonsilloliths were present over the soft palate. Less skillful dental radiologists might have more false-positive and false-negative cases of tonsilloliths on panoramic radiographs. Therefore, the particular radiological findings of tonsilloliths described above on panoramic radiographs should be highlighted in lectures for dental students and postgraduate dentists worldwide.

In the previous cases in children, only large tonsilloliths and clinical manifestations such as tonsillar abscess were described. However, even in children under 12 years old, there were some subjects (2.1%) with small tonsilloliths in the present study; thus, some tonsilloliths do occur in children. Since the presence of tonsilloliths was related to the occurrence of palatine abscess, pediatric dentists should be aware of the above.

As mentioned above, it was found that the prevalence of tonsilloliths was much greater than in previous reports [1–4]. The presence of tonsilloliths may be related to halitosis based on some previous reports [5, 7]. Therefore, some unknown causes of halitosis might be related to the presence of tonsilloliths. At the same time, we are planning to elucidate the relationship between the presence of tonsilloliths and halitosis in our next study. It is important to evaluate whether tonsilloliths are present on CT examination in patients with unknown causes of halitosis. If tonsilloliths are found incidentally on CT and/or panoramic radiographs in patients without a palatine abscess or halitosis, they should be followed regularly. However, if tonsilloliths are suspected as the causes of the palatine abscess or chronic halitosis, they should be removed [3].

The limitation of this study was that the sample size was small, and the data of subjects with diseases involving the oral and maxillofacial region were limited, except for the oropharyngeal area. Thus, the present results should be interpreted as being relevant to relatively healthy, active populations. A further limitation of this study was that only Japanese subjects were examined.

The present results suggest that the tonsillolith detection rate is relatively higher than suggested in previous reports. However, panoramic radiographs detected only a small percentage of palatine tonsilloliths. The differences mentioned above should be recognized. The low detection rates on panoramic radiographs might be related to the degree of calcification, the size, and the number of tonsilloliths.