Microscopic Features of Multifocal Fibrosing Thyroiditis

MFT is characterized by fibrotic scar-like lesions with jagged edges, containing a paucicellular fibrotic center surrounded by a cellular peripheral area with reactive-appearing follicular cell atypia and variable chronic inflammation (Fig. 1). Individual fibrotic foci display a zonal distribution of morphological alterations (Fig. 2).

Fig. 1

Multifocal fibrosing thyroiditis with fibrotic scar-like center surrounded by a cellular peripheral area

Fig. 2

Zonal distribution of pathologic changes in multifocal fibrosing thyroiditis

The central portion of the lesions is paucicellular and fibrotic (Fig. 3a and e), sometimes with fibroelastotic collagen similar to that seen in radial scar of the breast (Fig. 3e). The central area is also characterized by clusters of small follicles with scant colloid (Fig. 3b and f) occasionally displaying pseudo-infiltrative features (Fig. 3c and g), and scattered vessels sometimes with ectatic lumina and/or prominent walls (Fig. 3d and h).

Fig. 3

Zonal distribution of pathologic changes in multifocal fibrosing thyroiditis: central area. Dense paucicellular fibrosis (a) with fibroelastotic collagen similar to that seen in radial scar of the breast (e); clusters of small follicles (b, f); follicular cells with pseudoinfiltrative features (c, g); vessels with prominent walls (d, h). a, b, c, d: lower magnification. e, f, g, h: higher magnification

The peripheral portions show increased number of follicles lined by cells with reactive/regenerative atypia. Atypical features include nuclear enlargement and elongation with nuclear hyperchromasia (Fig. 4a and d) and chromatin clearing with minor irregularities of the nuclear contour (Fig. 4b and e). Nuclear grooves are rare. Within MFT foci there are no nuclear pseudoinclusions or psammoma bodies. Indeed, features diagnostic of PTC, including nuclear pseudoinclusions or psammoma bodies, were restricted to papillary carcinoma when this coexisted with MFT. Some of the follicles in the peripheral area may be elongated, with their main axis perpendicular to the MFT central fibrotic core (Fig. 4a and d). A predominantly chronic inflammatory infiltrate, including lymphocytes and histiocytes, sometimes with scattered multinucleated giant cells or granulomatous reaction, are also typically present at the periphery of the fibrotic core (Fig. 4c and f).

Fig. 4

Zonal distribution of pathologic changes in multifocal fibrosing thyroiditis: peripheral area. Increased number of follicles with reactive-appearing cytologic atypia (a, d, b, e) and predominantly chronic inflammation with granulomatous reaction (c, f). a, b, c: lower magnification. d, e, f: higher magnification

All our cases showed one or more foci with these features, that were identified in both conventional histology sections from Bologna cases and in the digital slides from Porto, with 100% concordance in the diagnoses between all pathologists involved in the study, including two (CE and GT) with a special interest in thyroid pathology.

The largest individual fibrotic foci often had a peripheral/subcapsular - as opposed to a central/intraparenchymal - location: 17/27 (63.0%) versus 10/27 (37.0%) (Table 1).

Table 1 Clinicopathologic features of multifocal fibrosing thyroiditis cases

Similar to what reported by Fellegara and Rosai [2], at low magnification we observed two main patterns of fibrosis: stellate (triangular in shape with the base situated parallel to the thyroid surface and one of the apices pointing to the center of the thyroid if the focus had superficial/subcapsular location) and band-like, the latter appearing as irregular foci of capsular thickening when superficial/subcapsular and as thickening of normal interlobular thyroid septa when the location was central/intraparenchymal. Stellate foci of fibrosis were in general the most common. Indeed, in individual MFT cases, the largest fibrotic focus was more often stellate (17/27, 63.0%) (Fig. 5a; Table 1) as opposed to band-like (10/27, 37.0%) (Fig. 5b; Table 1). In one case (BO-1), confluence of fibrotic foci resulted in a cirrhosis-like appearance (Fig. 5c), similar to case 8 of Fellegara and Rosai [2].

Fig. 5

Shape of fibrotic foci in multifocal fibrosing thyroiditis. Stellate (star-like) (a), band-like (b), cirrhosis-like (c)

Multiple fibrosclerotic foci were scattered within the thyroid gland in 22/27 MFT cases (81.5%), with a number of foci ranging from 2 to 17 per case (mean number: 4.6; median number: 4), and a mean diameter of 4 mm (range: 1 to 16 mm) (Tables 1 and 2). A single MFT focus was observed in 5/27 MFT cases (18.5%) with a mean diameter of 4.8 mm (range: 2 to 12 mm).

Table 2 Comparison of clinicopathologic features between routine cases and Dr. Rosai consultation series [2]

The site of MFT involvement within the thyroid gland was analyzed. Among cases with total thyroidectomy, the large majority had MFT in both lobes (10/17, 58.9%), 4/17 (23.5%) cases had MFT in the left lobe, 0/17 (0/17, 0%) in the right lobe, and 3/17 (17.6%) in the isthmus. Among specimens where only one lobe was resected, there were five cases with MFT in the right lobe, and five cases in the left lobe. Obviously bilaterality in lobectomy specimens cannot be evaluated. Thus, the overall majority of cases were bilateral, a relative minority consisted of cases that are at least in the left lobe (because only a left lobectomy was performed), followed by cases that were at least in the right lobe (because a right lobectomy was performed), and finally followed by cases that were in the isthmus (Table 1). The presence of only 4 left unilateral cases out of 17 total thyroidectomies was not statistically significant (P = 0.467).

Epidemiology of Multifocal Fibrosing Thyroiditis

In our multicentric retrospective study, we identified 27 cases of MFT among 1,736 consecutive and unselected thyroid resections, with an overall prevalence of 1.56% in the entire cohort. Of these, 22 cases were found over a seven-year period among the 1,393 resections retrieved from the Italian archive (1.58%), and 5 cases over a two-year period among the 343 (1.46%) resections from the Portuguese archive. The average yearly rate of MFT detection was 1.40% (range 0-2.80%) in the Italian thyroid resections, and 1.44% (range 1.28–1.60%) in the Portuguese ones. These figures correspond to an average yearly detection rate of 1.34% for the entire cohort.

Clinical Features of Multifocal Fibrosing Thyroiditis

MFT clinicopathologic features are shown in Table 1, with no significant difference between the Italian and Portuguese cases. There was a clear predominance of female patients: 23/27 (85.2%) were female and 4/27 (14.8%) male. However - in spite of a higher proportion of female patients in MFT - the correlation between MFT and patient sex in the 1,736 cohort population was not significant (P = 0.192), reflecting the general higher prevalence of thyroid pathology in the female sex. Age ranged from 34 to 69 years (mean: 54.4; median: 52.0) (Table 1).

Clinically, 7/27 patients (25.9%) were affected by autoimmune thyroiditis with persistently elevated anti-thyroglobulin antibodies in three. Four patients were clinically hypothyroid. Four of 27 patients (14.8%) suffered from hyperthyroidism, due in two cases to hyperfunctioning follicular adenoma, in one case to toxic multinodular goiter, and in one case to Graves disease.

Seventeen of 27 patients (63.0%) underwent total thyroidectomy, 5/27 (18.5%) right thyroid lobectomy, and 5/27 (18.5%) left thyroid lobectomy (Table 1).

In all cases but two, MFT was an incidental finding in thyroid glands resected for the presence of benign or malignant thyroid pathology. In 2/27 cases (7.4%) thyroid resection was performed for fibrotic MFT foci with worrisome ultrasound features and preoperative cytology that was indeterminate in the first case (BETHESDA III, lobectomy; BO-3) and suspicious (BETHESDA V, total thyroidectomy; BO-8) for papillary carcinoma in the second (Table 1). Malignant thyroid nodules - all papillary carcinomas - or other benign nodules were the reason for thyroid resection in 10/27 (37.0%) and in 12/27 (44.4%; Follicular adenoma: 3 thyroid resections; follicular nodular disease: 9 thyroid resections) cases, respectively. In the three remaining cases, resection was due to chronic lymphocytic thyroiditis (2/27 cases, 7.4%) and Graves disease (1/27 cases, 3.7%) (Table 1).

Thyroid Pathology Associated with Multifocal Fibrosing Thyroiditis

Meticulous histologic examination was performed to identify the spectrum of additional alterations present in thyroid specimens with MFT. Seventeen of 27 cases (63.0%) were associated with PTC, 10 cases being the reason for surgery (see above), and additional 7/27 cases (25.9%) being millimetric PTC found incidentally in specimens removed for benign thyroid pathology. Most PTC measured ≤ 10 mm (12/17, 70.6%), while the remaining cases (5/17, 29.4%) ranged from 13 to 25 mm; in both size groups, PTC were frequently multifocal (overall 9/17, 52.9%) (Table 1). The reason for surgery was benign thyroid pathology in the majority of cases (see above), but histologic analysis revealed a variety of benign thyroid alterations in 25/27 cases (92.6%), often overlapping with each other and with PTC (Table 1). Follicular nodular disease (17/27 cases; 62.9%) and chronic lymphocytic thyroiditis, including Hashimoto thyroiditis (14/27 cases; 51.8%), were the predominant benign histologic alterations in the cohort.

Comparison of Clinicopathologic Features with Dr. Rosai Consultation Series

We compared the clinical and pathologic features of our routine series with those of Dr. Rosai consultation series, which included 55 cases [2] (Table 2). The comparison of clinical features is summarized in Table 2a. Our routine cases developed in older patients (mean age: 54.4 versus 47.0; P = 0.006), but there was no statistical difference for sex, reason for surgery and type of surgery between the two studies (Table 2a). Comparison of MFT microscopic features is summarized in Table 2b. In our routine cases MFT tended to involve both lobes less frequently (Bilaterality: 58.8% versus 90.0%; P = 0.052) but the overall histologic features were no different (Table 2b). Comparison of MFT-associated thyroid pathology is summarized in Table 2c. In our routine cases MFT was more frequently associated with PTC (63.0% versus 31.6%; P = 0.022) or with benign follicular nodular disease (63.0% versus 39.4%; P = 0.080), although PTC dimensions and PTC multifocality were comparable between the two cases series (Table 2c).

Mutation status of Multifocal Fibrosing Thyroiditis

No genetic alteration was identified in any of the cases tested (15/27) for molecular analysis with a broad NGS panel designed to identify single nucleotide variants and small insertion/deletions in hot spot regions of genes relevant for thyroid tumorigenesis. Given the histologic similarity of MFT and PTC and their perceived link, an important reason to perform molecular analysis is to rule out BRAF p.V600E [8]. This was done on all cases by NGS and/or BRAF p.V600E immunohistochemistry, and in no case was BRAF p.V600E found: there was no positivity in any of the 20/27 cases, representative sections of which were stained with a BRAF p.V600E specific antibody. Neither BRAF p.V600E nor any other genetic alteration were identified in any of the 15/27 cases tested by NGS (Table 1).