Keywords:
Head and neck, Ultrasound, Biopsy, Pathology
Authors:
G. Hinton, S. Prowse
DOI:
10.26044/ranzcr2022/R-0080
Conclusion
Thyroid nodules are very common with up to 68% of adults reported to have a thyroid nodule (Guth et al 2009). FNA can be used to assess if these nodules are benign or malignant and guide further management if required.
Thyroid carcinoma is also common but not always clinically important and so choosing the correct patient to perform FNA is important, a previous systematic autopsy study found that 36% of 101 thyroid glands had occult papillary cancer (Harach HR et al 1985). The main criteria for biopsy based on ACR TIRADS is lesion size, as when malignancy is detected lesion size has been shown to be a significant prognostic factor, for example thyroid cancers <2cm have been shown to have an indolent course with a 99.9% 10-year survival rate (Nguyen XV et al 2013).
The reported non-diagnostic rate for thyroid FNA is quite varied ranging from 2-30%. The recommended diagnostic rate is 70% (Yong AW, Howelett D 2017). The main reasons for a non-diagnostic thyroid FNA include: insufficient material, sampling error or cytodiagnostic error (TL Hall et al 1989). Our audit has shown a slightly higher non-diagnostic rate (42%) than those previously reported including a previous South Australian audit which found a non-diagnostic rate of 13% (Goldfinch A, Prowse S 2018), although variability across institutions and study design limits direct comparison. There was no difference in technique in our audit in that almost all procedures were performed with a 25 gauge needle and the majority performed with the same number of passes. One factor that did show a difference in the diagnostic rate of the thyroid FNA was the level of experience of the doctor performing the procedure which is expected.
Technical difficulty of the procedure was not assessed by this audit with potential factors impacting diagnostic rate including depth of nodule, depth of soft tissue required to penetrate and patient anxiety/ movement. Unfortunately, due to the retrospective design of this audit and lack of routine documentation in reports these were unable to be assessed. Similarly, cytodiagnostic error was not formally assessed as this would require re-reading of the slides by separate pathologists and as this was not the primary aim of this audit, this was not explored.
Another observation from this audit is that a significant number of the thyroid FNA’s are performed which are not indicated based on the ACR TIRADS recommendations. The ACR TIRADs recommendations were established in 2017 by the ACR TIRADS committee to help guide the need for FNA or follow up of thyroid nodules given their prevalence. Prior to the implementation of the ACR TIRADS criteria there were a number of other consensus guidelines used which could be a contributing factor to the high number of FNA referrals not meeting current ACR TIRADS criteria for biopsy in this study. From the pathology reports of the FNA none of the non-indicated ultrasounds returned a malignant result.
Whilst this is a relatively small single centre audit some trends can be observed from the data collected. The overall diagnostic accuracy of thyroid FNAs in this audit were less than recommended and less than seen in similar studies suggesting that the change in practice may have contributed to this, although the lack of a control group is a clear limitation. The only factor observed to impact the diagnostic accuracy of the thyroid FNAs performed in this audit was level of experience of the performing clinician. There was no significant difference in procedure technique. The audit is unable to assess the underlying cause for the difference in diagnostic accuracy whether due to: insufficient sample, sampling error or poor slide preparation. We can hypothesize that consultants working in private practice are involved in slide preparation on a regular basis and hence this may be a contributing factor to the higher diagnostic rate in this cohort. Another observation from this audit is that there are many FNAs being performed which are not indicated based on the ACR TIRADS criteria; this could be addressed with referrers to determine other factors being considered when deciding on the indication for a thyroid FNA. Within the department it may be worthwhile giving the registrars and RMO’s a formal orientation into slide preparation with initial supervision from consultant radiologists to try to reduce this potential source of error going forward.