Clinical presentation: a 28-year-old woman was referred for an evaluation of a thyroid nodule. She was treated for hypothyroidism 14 years earlier.

Palpation: a firm nodule in the upper pole of the right lobe.

Hormonal investigation: indicated euthyroidism with TSH-level 0.81 mIU/L.

Ultrasonography: the thyroids were echonormal but contained around 30% of hypoechogenic areas. There was a hypoechogenic lesion in the upper ventral part of the right lobe. It presented irregular hyperechogenic granules and puzzle-like borders. The vascularization was not specific.

Cytological diagnosis: suspicion of an oxyphilic tumor.

Right lobectomy was performed. Histopathology disclosed Hashimoto's thyroiditis without any nodule.

Comment.

1. This was one of the most important case in my practice. I send to surgery a patient unnecessarily with the suspicion of tumor while histopathology disclosed that patient harbors not even a nodule. Considering the positive predictive value of thyroid cytology which ranges between 33 and 65%, this is a very common situation.

2. Moreover, at first sight we have nothing to do in this case and has to accept the well-known fact that the cost recognizing as much as possible carcinomas is to unnecessary operate some patients. The ultrasound pattern was suspicious and the cytological picture was also not calming. Firstly, there was an irregular, hypoechogenic lesion presenting microcalcifications, secondly, the smear contained almost exclusively oxyphilic cells.

3. The pivotal is to ask the right question. Which type of thyroid carcinoma could belong both the ultrasound and the cytological pattern ? The former might be the presentation of a papillary carcinoma while the cells on the smear could be arisen from oxyphilic variant of a follicular carcinoma. There was no cytological signs of a papillary carcinoma and there were no ultrasound signs of a follicular type tumor. The risk of a papillary carcinoma is certainly less than 2% if the nuclei lack inclusions while the risk of follicular carcinoma is less than 5% if the nodule lacks both halo sign and perinodular blood flow. Moreover, the underlying Hashimoto's thyroiditis explained the oxyphilic metaplasia.

4. This case and other similar cases resulted in a change of my way of thinking. If I met with a similar case today, I gave a common clinical-ultrasound-cytological diagnosis of "Hashimoto's thyroiditis. Hürthle-cell lesion with a less than 2% risk of carcinoma" and I recommended regular follow-up instead of surgery.