To our knowledge, this is the first survey of thyroid ultrasound characteristics conducted in patients with MNG in DICER1 syndrome who were all confirmed by genetic testing to have a pathologic germline. SAY1 mutation. We have presented the ultrasonographic features of MNG in DICER1 syndrome in descriptive terms. We took this approach because the ideal control group (children with MNG, who do not have germline DICER1 mutations) is not available and no publications exist for retrospective comparison. Indeed, it is likely that SAY1 is the main or only genetic cause of pure MNG in childhood, so it is difficult to compare DICER1 to other extremely rare syndromes, such as PTEN hamartoma tumor syndrome. Some studies have suggested that neuroblastoma (NB) may also be involved in this syndrome. Saskin et al.fifteen described the case of a 14-year-old female presenting with a multinodular goiter (MNG) and a collision tumor composed of NB and cystic nephroma (CN). She carried a deleterious germline mutation in exon 23 of SAY1 and harbored different somatic mutations in the CN and MNG. However, no second hit was found in the NB, leading to its status as a DICER1-related tumor being questioned. We have shown in Fig. 3a–c that such a family with MNG and the proband had a prior NB, but a SAY1 germline mutation was not present in this family.
The only one DICER1mut+ patient diagnosed in our center with thyroid microcarcinoma (PTCvF) had no classical ultrasound features of malignancy, but we cannot exclude that the other patients worldwide diagnosed with thyroid cancer may manifest such features.
Large cohorts, particularly those prospectively studied, may provide an answer to questions such as (a) at what age can thyroid nodular disease start in DICER1 syndrome; (b) how dynamic is the process; and (c) does it progress toward cancer regardless of whether it is initially benign or malignant, or do potential triggering factors need to be present? Larger patient series may allow for statistical analyzes and validation of our findings in the future.
The scant data on DTC in DICER1 syndrome may lead to questions regarding whether and when these patients should undergo thyroidectomy; moreover, what about the extent of surgery (total, near/total or not at all) if they still are euthyroid, have benign cytology from fine needle aspiration (category 2 based on Bethesda system)16, do not have symptoms from their goiter, or there is a very low risk for cancer? The most recent data on DICER1 syndrome in adolescent DTC suggest that germline mutations in SAY1 are more likely to lead to MNG than to DTC and that if DTC does occur in the context of biallelic SAY1 mutations, it is likely to be a low-risk tumor17. In a registry-based study, Khan et al.4 found that SAY1 carriers have a statistically significant, 16- to 18-fold increased risk of developing DTC and that this risk is increased to 24-fold after censoring of complete and partial thyroidectomies.
Among young individuals operated for benign MNG (nodular goitre, colloid nodular goitre, and hyperplastic nodular goitre) below age 25 years, 13% were found to possess a germline pathogenic variant in DICER118.
In two reports4.17 the authors stated that SAY1-associated thyroid cancer is not more invasive or less responsive to therapy than its non-SAY1-associated counterpart, which is supported by a more recent study17.
SAY1 Mutations in pediatric PTC are present at a frequency nearly 30 times that seen in adult PTC. Sequence of SAY1 identified pathogenic somatic variants in 10% PTCs, all of which lacked conventional alterations. germ line SAY1 pathogenic variants were identified in 20% of benign lesions. These data establish SAY1 as a common oncogenic driver in American Thyroid Association pediatric low-risk PTC and broaden our understanding of the molecular pathogenesis of pediatric PTCs17.
We agree with Khan and coworkers4 that SAY1 mutation carriers with a thyroid nodule should receive standard management before consideration for thyroid resection, including thyroid and neck ultrasonography, to assess evidence of bilateral thyroid disease and metastasis to cervical lymph nodes19.20. However, in our experience, these patients do not usually present with a solitary thyroid nodule but instead have at least three nodules; therefore, the diagnostic and therapeutic protocol should differ from that for solitary nodules. In our series, the decision whether to operate or not was based on clinical data, ie, the presence of a large multinodular goiter filling almost the entire volume of the gland. A fine needle biopsy ruled out thyroid cancer, although with multinodular lesions there is still uncertainty because not all nodules in these children are biopsied in such cases. At present, the literature indicates that the risk of cancer in children and adolescents is low, but preoperative biopsy does not detect all thyroid cancers. Another facet is that in the context of DICER1 syndrome, it is extremely likely that a partial thyroidectomy will need to be followed by a completion thyroidectomy at some point. In addition, childhood- and adolescent-onset poorly differentiated thyroid carcinoma (PDTC) has been shown to be associated with SAY1 mutations and may herald DICER1 syndrome in some patients. Furthermore, their clinically aggressive behavior contrasts sharply with the indolent nature of the great majority of thyroid tumors with SAY1 mutations reported to datetwenty-one. Thus, while most lesions will follow a benign course, total thyroidectomy is the procedure of choice for the above-mentioned reasons.
Total thyroidectomy as the procedure of choice is recommended for the above-mentioned reason, since it avoids further glandular proliferation of an already significantly enlarged gland with features of multinodular goiter and nodular disease recurrence in the case of partial thyroidectomy and allows to diagnose cancer at early stage.
Clinical and ultrasonographic follow-up on 6-month intervals or at least on annual basis is mandatory in non-operated patients with DICER1-related MNG in order to detect early a rare but potentially possible transformation into the suspicious lesion that need biopsy and more radical treatment.
In terms of management of other mendelian forms of MNG, a retrospective review of prospectively accrued PTEN hamartoma tumor syndrome (PHTS) patients suggests stratifying surveillance intervals based on thyroid ultrasound result, and support extending surveillance intervals in PHTS patients without nodules on ultrasound to 3– 5 years, and patients with clinically nonactionable nodules to 2–3 years, which is in contrast to the current recommendation of annual ultrasounds. This change in practice would decrease the burden of frequent ultrasounds, especially in young children and adolescents that are more likely to have a normal or nonactionable ultrasound result.24.
Long-term prospective follow-up of patients with DICER1 syndrome who remain thyroid disease-free is ongoing and may shed a new light on active surveillance in SAY1mut+ patients to propose the most accurate frequency of controlled ultrasound examinations.
Clinicians should consider in the diagnostic workup that other hereditary disorders/syndromes (Suppl. Table 1) may also manifest with multinodular goiter, but based on our experience, nodules in patients with DICER1 syndrome, particularly those with benign histopathology and those presented in this paper , have a unique and distinguishable pattern. Other etiologies may contribute in MNG pathogenesis, however genetic causes of euthyroid MNG in childhood are all likely to be much less prevalent than DICER1 syndrome.
Euthyroid MNG before 30 years strongly suggests a hereditary cause. DICER1 is in our opinion the dominant candidate in such patients particularly if the ultrasound features we report here are present. The diagnostic protocol should also consider investigation for DICER1 large deletions, either by MLPA22 or as part of a larger NGS panel (including genes in Supplementary Table 1) that is capable of detecting copy number variants.
We hope that our US criteria of MNG in DICER1 syndrome, when compared with a classic ultrasonographic presentation of thyroid cancer, will shed new light on which children and adults with MNG should be considered for SAY1 genetic testing. These criteria could also be useful in determining how likely it is that malignancy will be present. In this vein, the lack of suspicious cervical lymph nodes seems to be an important and adjuvant ultrasonographic feature in evaluating MNG patients. Fine needle aspiration of thyroid nodules is a superior tool for detecting thyroid cancer to optimize the surgical extent and improve the final outcome. We also agree with Khan and coworkers4 that there are no data to support prophylactic thyroidectomy in those with DICER1 syndrome, unlike other hereditary syndromes affecting the thyroid, such as multiple endocrine neoplasia type 223. However, if surgery is being performed as definitive treatment, then total thyroidectomy should be advocated, despite the almost universal absence of US features of thyroid malignancy in SAY1-mutated MNGs. There is still no answer to the question of whether adjuvant therapy with low doses of radioiodine is warranted for DICER1 syndrome-associated DTC, but based on the current data, we favor a “wait-and-see” approach with annual clinical, hormonal and ultrasonographic evaluation until long-term data are available. Another related issue is the controversial morphological/histopathological presentation of MNG in DICER1 syndrome, given the tendency to diagnose papillary thyroid carcinoma in borderline cases where papillomatous hyperplasia can mimic papillary thyroid carcinoma. We suggest that historically diagnosed PTC with DICER1-related pathological features of their MNG should be re-examined by pathologists with experience in DICER1-related tissue abnormalities.