The whole 41 patient cohort contained 22 females and 19 males (Table 1), ranging from 18–87 years (median 63 years; average 58.6 years). Patients reported symptoms from 1-480 months (median 8 months; average 69 months) in the HGDFCDTC group and 1-355 months (median 6 months; average 46 months) in the PDTC group. Of note, one HGDFCDTC case presented incidentally during imaging evaluation for a different reason, while four PDTC patients had incidental tumors identified. All patients in the HGDFCDTC had preoperative fine needle aspiration (FNA), although with a range of up to five FNAs performed: five Bethesda category II; 5 category III; four category IV; and three category V; five cases had molecular investigation, including four with multiple mutations (TERT promoter mutations (n = 2); PTEN (n = 2), NRAS (n = 1), PIK3CA, PPP2R1A, DAXX, NF1, CDKN2C, ARID1A, PIK3R1, RUNX1T1), and one PAX8::PPARγ. Twenty-three patients in the PDTC group had FNAs performed, with up to five FNAs performed: three category II; six category III; 10 category IV; three category V; and one category VI; 13 cases had molecular investigation, including three with no mutations identified; four with multiple mutations (NRAS (n = 4); TERT promoter mutations (n = 2); KAT6B, MUTYH, PALB2), NRAS p.Q61K (n = 3), and one each WHSC1L1::NUTM1, PTEN deletion, and PAX8::PPARγ. Of note, one tumor in the HGDFCDTC category and three in the PDTC category lacked capsular or lymphovascular invasion, but met all other histologic criteria for inclusion, and were thus included as a subset of cases that may lack invasion and still qualify for inclusion in the category (11{Rivera, 2010 #6012)}. All cases presented with a nodule/mass, although multinodular goiter was noted in six HGDFCDTC and nine PDTC cases. Overall, eight of 41 patients (19.5%) had metastatic disease at presentation (lymph node = 5; lung = 3; bone and mediastinum = 1 each). Using the criteria established above, tumors were separated into the two categories for further evaluation.
Pathologic Features
Macroscopic findings
The HGDFCDTC group included nine PTCs (four solid; three classical; two follicular subtype) and eight oncocytic FTCs. The PTCs ranged from 3.8–13.5 cm (median 6 cm; average 6.6 cm), with well-defined capsules. Five presented as a single tumor and four were part of multifocal/bilateral tumors. Eight were widely invasive, all with capsular invasion, seven with lymphatic invasion, six with vascular invasion, while one tumor lacked invasion. The OFTCs ranged from 2.5-7 cm (median 4.8 cm; average 4.8 cm), with well-defined capsules. All were identified as single tumors without other neoplasms identified. Seven of eight were widely invasive, with capsular and vascular invasion identified, while only one tumor had fewer than 4 foci each of capsular and vascular invasion.
The PDTC cases ranged from 1.9 to 11 cm (median 6.9 cm; average 6.5 cm), with well-formed capsules identified. Eleven presented as single tumors, with the remaining 13 identified as part of multifocal/bilateral disease. All of the additional tumors were papillary carcinomas, although an additional patient had a metastatic neuroendocrine carcinoma, small cell type to the thyroid gland. Based on these findings, there was no statistically significant difference in size between the tumors, although there was trend to the PDTC being larger. Multifocality was more common with the PDTC category (54 vs. 24%).
Microscopic findings
HGDFCDTC
The tumors demonstrated a well-developed fibrous connective tissue capsule with smooth muscle-walled vessels in the fibrosis, the latter a finding support a genuine capsule. One tumor lacked invasion (a 3.8 cm unifocal solid PTC; Ki-67 labelling index 8.3%; alive, no evidence of disease at 17.4 months), but all of the others showed capsular invasion (n = 16), lymphatic invasion (n = 8), and/or vascular invasion (n = 14). One oncocytic follicular carcinoma was minimally invasive, four tumors were invasive, and the remaining 11 tumors were widely invasive. By definition, the latter tumors showed either > 20 foci of capsular and/or > 20 foci of either lymphatic or vascular invasion. For example, a widely invasive PTC had > 10 but ≤ 20 foci of capsular invasion and > 10 but ≤ 20 foci of lymphatic invasion without vascular invasion (but a total of > 20 foci of invasion) while an OFTC showed > 20 foci of capsular invasion and > 20 foci of both lymphatic and vascular invasion (a total of > 40 foci of invasion). Thus, widely invasive could be reached by several factors. Extrathyroidal extension (by current definitions) was identified in four tumors (two classical PTC and two OFTCs). Tumor was identified on the inked margins in nine tumors. Lymph node metastases were identified in two patients at the time of initial presentation. Based on these findings, there were seven group I, seven group II, two group III, and one group IVA staged tumors.
Histologically, the classical morphologic features of the underlying tumor type were present, as generally accepted for PTC (classical; invasive follicular and solid subtypes) and OFTC (Fig. 1). For clarification, the solid subtype of PTC frequently shows overlap with PDTC with the solid, trabecular or nested growth pattern, but shows classical nuclear features of PTC, not seen in PDTC; further, tumor necrosis and increased mitoses are not present in the solid subtype of PTC while they are in the HGDFCDTC and PDTC. OFTC is composed by at least 75% oncocytically altered follicular-cell derived cells lacking nuclear features of PTC and high-grade features (Fig. 1). Importantly, areas of infarction or degeneration due to outgrowing its blood supply or after needle biopsies (FNA or core needle) must be carefully examined, with only genuine, tumor comedonecrosis as previously defined, used to classify true tumor necrosis.
Mitoses were sought in the regions of highest cell density, around areas of tumor necrosis, or at the tumor periphery where fixation was optimal. Two mm2 were counted on three different slides and the average recorded (see Table I; Fig. 1). It is important to note there were no significant differences when counting multiple different 2 mm2 regions, and for practical application, a single 2 mm2 region is sufficient, using tumor cellularity as a guide. Mitotic counts ranged from 1 to 14/2 mm2, with a median of 5 (SD 3.4) and average of 6. The number of cells in 2 mm2 ranged from 4,024 to 11,846, with a median of 8,708 (average 8119), with the number of Ki-67 positive nuclei ranging from 278 to 1874, with a median of 684 (average 775), yielding a Ki-67 labeling index from 2.3 to 19.6% with a median of 8.3% (SD 4%; average 9.4%).
In the background, four cases had another tumor identified in the same lobe as well as in the completion thyroidectomy performed in three of the cases. They were classical (including microscopic) and follicular subtype. Other thyroid gland disease was noted: multinodular hyperplasia (adenomatoid nodules; n = 9, 53), chronic lymphocytic thyroiditis (n = 3, 18%), unremarkable parathyroid gland tissue (n = 2, 12%), and included benign perithyroidal lymph nodes (n = 5, 29%).
PDTC
The tumors demonstrated a well-developed fibrous connective tissue capsule with smooth muscle-walled vessels in the fibrosis. Three tumors lacked invasion (one with multifocal disease; two with unifocal tumors; all alive with no evidence of disease, median 19.0 months). The remaining 21 all showed capsular invasion, with lymphatic invasion (n = 15), and/or vascular invasion (n = 16). Five tumors were invasive, and the remaining 16 tumors were widely invasive (as previously described). There was extrathyroidal extension in nine tumors, while tumor was identified on the inked margins in nine tumors (six of these cases showed both features). Lymph node metastases were identified in three patients at the time of initial presentation. Based on these findings, there were 10 group I, 12 group II, one group III, and one group IVA staged tumors.
Histologically, tumor cells were arranged in insular, solid and trabecular architecture, with follicular structures noted in many areas to confirm follicle-cell derivation (Fig. 2). Drop-like colloid was noted in the center of follicles, as an opacified eosinophilic substance, highlighted by thyroglobulin (Fig. 2), although not a required immunohistochemistry evaluation for the tumor classification. The insular growth had nests surrounded by thin fibrovascular septa, with a separation or cleft between the stroma and neoplastic cells. This separation created a patulous to open or fenestrated appearance in these areas (Fig. 2). Trabeculae were wider ribbons of neoplastic cells. The tumors tend to be quite monotonous, with the neoplastic cells relatively small with a very high nuclear to cytoplasmic ratio (Fig. 2). There was usually scant to almost absent cytoplasm in most of the tumors. The nuclei appeared raisinoid, with dense, heavy nuclear chromatin, and occasional small nucleoli. Nuclear convolutions, folds or grooves could be seen (Fig. 2), but the composite nuclear features of PTC were absent. Isolated nuclei were more vesicular, but this was not the dominant finding. Intranuclear cytoplasmic inclusions were absent. Mucinous, signet-ring and rhabdoid cells were not identified. These features were not isolated, but composed more than 50% of the volume of the tumor, and in many cases the only pattern or histomorphology present. Again, based on previous descriptions, tumor necrosis and average mitotic counts were recorded (see Table I). Mitotic counts ranged from 2 to 39/2 mm2, with a median of 6 (SD 10.6) and average of 11. The number of cells in 2 mm2 ranged from 4,962 to 14,200, with a median of 10,277 (SD 3182; average 9911), with the number of Ki-67 positive nuclei ranging from 26 to 1,924, with a median of 620 (SD 734, average 833), yielding a Ki-67 labelling index from 0.3 to 27.2% with a median of 6.9% (SD 7%; average 8.4%; Fig. 3).
In the background, three cases had multifocal (same lobe) PTC, while 10 cases had bilateral (contralateral lobe) tumors: eight PTC, one additional PDTC, and one metastatic small cell carcinoma to the thyroid gland. The PTCs were classical (n = 2), microscopic classical (n = 7), and follicular subtype with invasion (n = 1)(more than eight total, as several microscopic PTCs may have been present). Other thyroid gland disease was noted: multinodular hyperplasia (adenomatoid nodules; n = 15, 63%), chronic lymphocytic thyroiditis (n = 7, 29%), diffuse hyperplasia (n = 1, 4%), unremarkable parathyroid gland tissue (n = 2, 8%), parathyroid gland adenoma (n = 1, 4%), and included benign perithyroidal lymph nodes (n = 5, 21%).
Clinical Treatment And Patient Outcome
Patients were managed by surgery only (n = 6); radioablative iodine in 28; external beam radiation in eight; and chemotherapy in five, with various combinations of the latter therapies in several patients.
Of those with surgery only, all six are without disease, five alive and one dead (median 4.0 months, average 3.7 months), but only followed for a short period. Surgery and radioablative iodine was used in 29 patients, but 6 of these also had other therapies; thus only 24 patients were managed by just surgery and radioablative iodine: 20 of 24 were without evidence of disease (alive = 19; dead = 1; median 34.3 months; average 37.1 months); and four were dead with disease (median 28.0 months; average 26.5 months). Those managed with follow-up radioactive iodine ablation treatment were treated by a range 29—212 µCi 131I (median 153; average 153.9). External beam radiation was used in eight patients. Chemotherapy was employed in five patients, although chemotherapy was also used to treat a separate pancreatic adenocarcinoma and widely metastatic small cell carcinoma, but without a specific regimen for thyroid cancer.
HGDFCDTC
In the HGDFCDTC group, overall follow-up was a median of 23.9 months (average 27.0 months), with 10 alive without evidence of disease (median 19.7 months; average 22.5 months), one dead without evidence of disease (24.8 months), four alive with metastatic disease (median 25.8 months; average 37.8 months), and two dead with metastatic disease (median and average 29.4 months). Thus overall, there were six patients with metastatic disease at last follow-up (median 25.8 months, average 35.0 months) and 11 patients without metastatic disease (median 21.2 months, average 22.7 months); overall 41.2% (7/17) had developed metastatic disease (one patient has no evidence of disease after treatment for lymph node metastases at presentation).
The overall survival of patients with invasive tumors showed all five alive (median 38.2 months), but one with metastatic disease (median 42.5 months), compared to widely invasive tumors showing six alive and one dead without evidence of disease (median 29.9 months) and five with metastatic disease (two alive and 3 dead; median 38.0 months); 41.7% with widely invasive tumors developed metastatic disease.
Three patients (17.6% had metastatic disease at presentation, two to lymph nodes and one to lymph nodes and lung. Two patients were dead with disease (median and average, 29.4 months) and one was alive without disease (21.2 months).
Four patients were managed by surgery alone: one developed lung metastatic disease 28 months after surgery treated with radioablative iodine at that time, and he is alive with metastatic disease at 27.6 months (he also has metastatic prostate adenocarcinoma being separately managed); the other three are alive without disease (median 3.2 months). One patient had surgery and external beam radiation and is alive without evidence of disease at 33.0 months. One patient had surgery without radioablative iodine, but had metastatic lung disease at the time of presentation, treated with chemotherapy, and died with disease 6.8 months after surgery. Eleven patients had surgery followed by immediate radioablative iodine: one patient had metastatic disease at presentation, with subsequent metastatic disease to lymph nodes and lung, with various chemotherapy regimens employed, he died with disease at 52 months; one developed metastatic disease to lymph nodes 79 months after surgery, and is alive with disease at 87.8 months after external bean radiation and additional radioablative iodine; one developed lung metastatic disease 4 months after surgery and received chemotherapy and radioablative iodine and is alive with disease at 11.6 months; one developed metastatic disease to the lungs 16 months after surgery, managed with external bean radiation and chemotherapy and is alive with disease at 23.9 months; the remaining seven patients remain without metastatic disease (one died of renal failure; six are alive), median 24.8 months (average 29.6 months).
Overall, six patients were < 55 years at presentation: one alive with metastatic disease (23.9 months), and the remaining five alive without evidence of disease (median 18.3 months, average 24.5 months); 16.6% developed metastatic disease. Eleven were ≥ 55 years at presentation: three alive with metastatic disease (median 27.7 months; average 42.4 months), two dead with metastatic disease (median and average 29.4 months); and six alive or dead without metastatic disease (median 28.8 months; average 21.2 months); 45.6% developed metastatic disease. Older age at presentation seems to influence overall outcome.
Overall, nine patients were female: one died with metastatic disease (6.8 months), and the remaining eight are alive without evidence of disease (median 11.9 months, average 14.8 months); 11.1% developed metastatic disease. Eight patients were male: three alive with metastatic disease (median 27.7 months; average 42.4 months), two dead with metastatic disease (median and average 29.4 months), and six alive or dead without metastatic disease (median 28.8 months; average 21.2 months); 45.6% developed metastatic disease.
Tumors were categorized using pT criteria: no pT1a or T1b tumors; three pT2 tumors, with one male alive with disease (11.6 months) and two patients alive without disease (median and average 18.1 months); 33% developed metastatic disease. The remaining 14 patients all had pT3a/b tumors: nine without disease at last follow-up (dead = 1; alive = 8; median 21.2 months; average 23.7 months), although one had metastatic lymph node disease at presentation; five with metastatic disease (dead = 2; alive = 3; median 27.7 months; average 39.6 months); 42.9% developed metastatic disease. While this was a trend, size did not seem to influence outcome, as the patient with the smallest tumor died of metastatic disease in 12 months, and the patient with the largest tumor is alive with metastatic disease at 42.5 months.
Overall, seven patients had Group I stage: one male is alive with metastatic disease (23.9 months), and the remaining six are alive without evidence of disease, although one had lymph node metastasis at presentation (median 19.7 months, average 25.9 months); 28.6% developed metastatic disease. Seven patients had Group II stage: one male is alive with metastatic disease (27.6 months) and one is dead with metastatic disease (52.0 months), while the remaining five are without evidence of disease (alive = 4; dead = 1; median 24.8 months; average 18.9 months); 28.6% developed metastatic disease. Three patients had Group III/IVA stage: two males are alive with metastatic disease (median and average 49.7 months), and one female is dead with metastatic disease (6.8 months); 100% developed metastatic disease.
Related to stage, but separate is the presence of extrathyroidal extension, defined by gross strap muscle invasion (pT3b) or gross extrathyroidal extension into other neck organs, subcutaneous tissues, prevertebral fascia, or named nerves and/or vessels identified by imaging findings, intraoperatively by the surgeon, or macroscopically by the gross prosector. Using these criteria, four patients had extrathyroidal extension: one female was alive without metastatic disease (1.8 months); three with metastatic disease: two alive (median and average 49.7 months), and one dead with disease (52 months); 75% developed metastatic disease.
An attempt was made to risk stratify based on Ki-67 labelling index, using 5% incremental cutoffs (0-4.9%; 5-9.9%; 10-14.9%; 15-19.9%). In the < 5% Ki-67 labelling index, one of three (33%) patients developed metastatic disease; in the 5-9.9% group, four of six (67%) patients developed metastatic disease; in the 10-14.9% group, two of six (33%) patients developed metastatic disease; and in the 15-19.9% group, one of two (50%) patients developed metastatic disease. Thus, no Ki-67 labelling index can meaningfully predict who may develop metastatic disease.
PDTC
In the PDTC group, overall follow-up was a median of 27.7 months (average 29.5 months), with 17 alive without evidence of disease (median 34.0 months; average 34.3 months), one dead without evidence of disease (4.7 months), two alive with metastatic disease (median and average 7.3 months), and four dead with metastatic disease (median 28.0 months; average 26.5 months). Thus overall, there were six patients with metastatic disease at last follow-up (median 15.5 months; average 20.1 months) and 18 patients without metastatic disease (median 33.4 months, average 32.6 months); overall 29.2% had developed metastatic disease (two patients had no evidence of disease after treatment for lymph node metastases at presentation).
The overall survival of patients without invasion (n = 3): all alive without disease (median 7 months; 20.3 average months); invasive tumors showed all five were alive without evidence of disease (median 35.0 months; average 34.2 months); compared to 16 widely invasive tumors: nine alive and one dead without evidence of disease (median 28.3 months; average 35.4 months) and six with metastatic disease (two alive and four dead; median 15.5 months; average 20.1 months); 43.8% with widely invasive tumors developed metastatic disease.
Five patients (20.8%) had metastatic disease at presentation, three to lymph nodes, two to lung, and one each to bone and mediastinum. Three patients had metastatic disease at last follow-up (alive = 1; dead = 2; median 4.9 months; average 18.3 months), and two were alive without metastatic disease (median and average, 8.7 months).
Three patients were managed by surgery alone: all are without evidence of disease but only followed for a short duration, with one dead of metastatic small cell carcinoma (4.7 months): median 4.7 months, average 4.2 months. Seventeen patients had surgery followed by immediate radioablative iodine: three patients had metastatic disease at presentation (two patients with lymph node metastases; one with bone and lung metastases); two patients developed metastatic disease to lymph nodes, liver and bone 15 and 19 months after presentation. As a group, 13 were alive without metastatic disease (median 41.5 months; average 41.2 months); four were dead with metastatic disease (median 28.0 months; average 26.5 months); 29.4% developed metastatic disease. Three patients had surgery without radioablative iodine, but had external bean radiation (n = 3) in addition to chemotherapy (n = 1): two are alive without metastatic disease (median and average, 19.6 months), and one is alive with metastatic disease (4.8 months); 67% developed metastatic disease. The remaining patient had surgery, radioablative iodine and external beam radiation, and is alive with metastatic disease (9.8 months).
Overall, 11 patients were < 55 years at presentation: one alive with metastatic disease (4.8 months), eight alive without evidence of disease (median 24.7 months, average 32.7 months), and two dead with disease (median and average 33.2 months); 36.4% developed metastatic disease. Thirteen were ≥ 55 years at presentation: one alive with metastatic disease (9.8 months), two dead with metastatic disease (median and average 19.8 months); and 10 alive or dead without metastatic disease (median 34.5 months; average 32.6 months); 30.8% developed metastatic disease. Age at presentation did not seem to influence overall outcome.
Overall, 13 patients were female: 2 had metastatic disease at presentation; one patient died with metastatic disease (4.9 months), and the remaining 12 are without evidence of disease (alive = 11, dead = 1; median 33.4 months, average 32.2 months); 15.4% developed metastatic disease. Eleven patients were male: two were alive with metastatic disease (median and average 7.3 months), three were dead with metastatic disease (median 34.8 months; average 33.7 months), and six alive or dead without metastatic disease (median 36.4 months; average 33.3 months); 54.5% developed metastatic disease. Males were more likely to develop metastatic disease.
Tumors were categorized using pT criteria: one female T1b tumor alive without disease (102.9 months); four pT2 tumors, all without disease (alive = 3; dead = 1; median 22.1 months; average 29.5 months); 0% developed metastatic disease. The remaining 19 patients all had pT3a/b tumors: two alive with metastatic disease (median and average 7.3 months), four dead with metastatic disease (median 28.0 months; average 26.5 months), and 13 alive without disease (median 34.0 months; average 28.1 months), although two had metastatic lymph node disease at presentation; 42.1% developed metastatic disease. Advanced size increased the risk of developing metastatic disease.
Overall, 10 patients had Group I stage: one male is dead with metastatic disease (21.3 months), and the remaining nine are without evidence of disease, although one had lymph node metastasis at presentation (median 16.5 months, average 29.5 months); 20% developed metastatic disease. Twelve patients had Group II stage: one male is alive with metastatic disease (4.8 months), three are dead with metastatic disease (median 34.8 months; average 28.2 months), while the remaining eight are alive without evidence of disease (median 42.5 months; average 35.9 months); 41.7% developed metastatic disease. Two patients had Group III/IVA stage: one male is alive with metastatic disease (9.8 months), and one female is alive without metastatic disease (34.0 months); 50% developed metastatic disease. Advanced stage is associated with a higher risk of developing metastatic disease.
Related to stage, but separate is the presence of extrathyroidal extension (see above). Nine patients had extrathyroidal extension: five are alive without metastatic disease (median 34.0 months; average 44.7 months), and four had metastatic disease at last follow-up (alive = 2; dead = 2; median 7.3 months; average 10.2 months); 67% developed metastatic disease.
An attempt was made to risk stratify based on Ki-67 labelling index, using 5% incremental cutoffs (0-4.9%; 5-9.9%; 10-14.9%; 15-19.9%). In the < 5% Ki-67 labelling index, two of seven (29%) patients developed metastatic disease; in the 5-9.9% group, three of 10 (33%) patients developed metastatic disease; in the 10-14.9% group, one of three (33%) patients developed metastatic disease; and in the 15-19.9% group, two of four (50%) patients developed metastatic disease. Thus, no Ki-67 labelling index can meaningfully predict who may develop metastatic disease, although there is a trend of development metastatic disease with a higher labelling index.