How many incidental findings do we detect on orthodontic CBCTs?
There is currently a debate on the risk/benefit balance of CBCT images for orthodontic patients. Some people suggest that the benefits of using CBCT images outweigh the risks. In contrast, others feel that we cannot justify the increased radiation dose. This new paper looked at the number of incidental findings detected on orthodontic CBCT views.
We all know that CBCT images provide much more information than conventional 2D views. However, while this information appears to be valuable, we also realise that we obtain this at the cost of increased radiation exposure. Some proponents of CBCT state that one advantage of using these techniques is that we may find Incidental Findings (IFs) that we would miss with conventional 2D methods. However, there is little large scale research that has looked at this area.
This new study provides us with some helpful information.
A team from Korea and the USA did this study. Nature Scientific Reports published the paper. It is open access.
Jin‑Young Choi et al. Nature: Scientific Reports.
I have not reviewed a paper published in Nature before. It is challenging to get a report published in these journals, and I am not aware of such a high impact journal publishing an orthodontic article.
What did they ask?
They did this study to answer two questions:
“What is the rate of incidental findings of craniofacial disease or abnormal structures from 3D CT images that were taken for orthodontic diagnosis without other symptoms”? And
“What is the detection accuracy of 2D radiographs compared to 3D CBCT images”?
What did they do?
They did a large retrospective study of case records of patients who had attended a dental hospital orthodontic clinic from January 2010 to July 2019. They had taken the CBCT images for orthodontic purposes only. Notably, the images had a wide field of view.
They found a total of 1020 images taken for 1020 patients. They also collected 2 D images from the patients.
Two oral and maxillofacial radiologists then evaluated the images. They classified the IFs into five categories. These were
- Maxillary sinus
- Nasal Cavity
- Soft tissue calcification
They classified these as the “Gold Standard”.
The same two examiners reviewed the 2D images and compared the Incidental Findings to the CBCT “gold standards”. This design gave them sufficient information to calculate the sensitivity/specificity, positive/negative predicted value and accuracy of the 2D images.
What did they find?
Their most significant finding was that they found at least one IF in 709 (69%) images. The most common of these were nasal cavity problems, for example, nasal septum deviation. The mean age of the patients with IFs was 22.3 years old, and the mean age for those without was 20.1 years old.
The highest accuracy for the 2D radiographs was 85% for soft tissue calcification. The lowest was 62.2% for nasal cavity findings.
In their discussion, the authors pointed out that these findings were similar to other studies. However, their sample was the largest that any researchers had studied to date. They also pointed out that the FOV of their sample was very large and evaluated the entire craniofacial area.
Importantly, they did not find any malignancies or life-threatening pathologies in the CBCT images.
Their conclusions were;
“69% of the CBCT views contained an incidental finding. As a result, clinicians should be responsible for carefully investigating their findings”. and
“The accuracy of detecting IFs on 2D views was not high”.
What did I think?
This paper is very interesting and relevant to clinical practice. I was surprised at the high number of IFs that they detected in their population of patients. I was even more surprised that other authors had found similar levels. However, and I may be being a bit cynical here. What I want to know is how many of these IFs warranted onward referral and treatment. Importantly, the authors mention that none of the patients had a life-threatening severe IF. Indeed, we may consider these IFs to be”normal” and expected.
These findings bring me back to the previous discussion. If we feel that the incidental findings are minor and not of clinical significance, we cannot justify taking higher radiation dose images. Paradoxically, with such a high level of IFs, we also need to ensure that we are carefully viewing the 3D images. As a result, we must be proficient in 3D image analysis or send our CBCTs to radiologists. This dilemma is real, and I know that I would err on the side of caution and send my CBCTs for expert evaluation. This practice was routine in the NHS Children’s Hospital that I used to work in.
This paper deserves a debate. This subject is undoubtedly controversial. However, I am finding it difficult to interpret its findings clinically because of the absence of information on the action taken following the identification of the IFs. Nevertheless, if I were still working clinically, I could not justify taking CBCTs in the search for incidental findings. Furthermore, I would get a radiologist to report the images. Some may think that this is too cautious, so let’s have a debate in the comments.
Emeritus Professor of Orthodontics, University of Manchester, UK.