Fax: (415) 885-7617
Follicular thyroid carcinoma: Histology and prognosis
Article first published online: 13 FEB 2004
Copyright © 2004 American Cancer Society
Volume 100, Issue 6, pages 1123–1129, 15 March 2004
How to Cite
D'Avanzo, A., Treseler, P., Ituarte, P. H. G., Wong, M., Streja, L., Greenspan, F. S., Siperstein, A. E., Duh, Q.-Y. and Clark, O. H. (2004), Follicular thyroid carcinoma: Histology and prognosis. Cancer, 100: 1123–1129. doi: 10.1002/cncr.20081
- Issue published online: 5 MAR 2004
- Article first published online: 13 FEB 2004
- Manuscript Accepted: 17 DEC 2003
- Manuscript Revised: 21 NOV 2003
- Manuscript Received: 26 AUG 2003
- Friends of Endocrine Surgery
- Gerald Heller Family Foundation
- Albert Clark Family Foundation
- Sanford and Helen Diller Foundation
- Mount Zion Health Systems
- follicular thyroid carcinoma;
- capsular and angioinvasion;
- histologic classification of follicular thyroid carcinoma
Follicular thyroid carcinoma (FTC) is the second most common thyroid malignancy after papillary thyroid carcinoma. The authors studied the clinical course of 132 patients with FTC to determine whether there was a direct relation between the histologic degree of invasion, tumor recurrence, and patient survival.
The 132 patients in the study population underwent 182 thyroid carcinoma–related operations, and their mean follow-up was 7.5 years (median:,6 years; range, 0–39 years). The following criteria were used to define malignant follicular neoplasms: 1) minimally invasive, tumor invasion through the entire thickness of the tumor capsule; 2) moderately invasive, tumor with angioinvasion (with or without capsular invasion); and 3) widely invasive, broad area or areas of transcapsular invasion of thyroid and extrathyroidal tissue. Forty-five of 119 patients (37.8%) presented with minimally invasive FTC (capsular invasion only), 50 patients (42%) presented with moderately invasive FTC (angioinvasion with or without capsular invasion), and 24 patients (20%) presented with widely invasive FTC. At presentation, 12 patients (9%) had distant metastases, and 8 patients (6%) had lymph node metastases.
Excluding 12 patients who presented with distant metastases, 21 patients (16%) developed recurrent metastases 6 months after their initial treatment. Among 45 patients with capsular invasion only, 6 patients (13%) developed recurrent or persistent disease, and 5 patients (11%) died. Of the 50 patients who had angioinvasion with or without capsular invasion, 10 patients (20%) developed recurrent or persistent disease, and 7 patients (14%) died. Patients who had angioinvasion with or without capsular invasion had a less favorable prognosis compared with patients who had capsular invasion only (P < 0.0001). Among patients who had widely invasive FTC, 9 of 24 patients (38%) developed recurrent disease, and 8 patients (33%) died; in addition, 7 of the other 24 patients (29%) had persistent disease and died. The overall death rate for patients with widely invasive FTC was 62%. Patients with persistent disease had a poorer prognosis compared with patients who had recurrent disease (P < 0.0001). Twenty-eight patients (21%) in the entire group died of FTC.
In the current retrospective investigation, the authors demonstrate that patients with minimally invasive FTC (capsular invasion only) had a slightly better survival rate at 5 years (98%) compared with patients who had angioinvasion with or without capsular invasion (80%) and had better survival compared with patients who had widely invasive FTC (38%). Other (but not all) reports in the literature support the findings that FTC with angioinvasion is more aggressive than FTC with only capsular invasion yet is less aggressive than widely invasive FTC. The authors conclude that FTC no longer should be classified as either minimally invasive or widely invasive; rather, they recommend classifying FTC as minimally invasive, moderately invasive, or widely invasive, because prognosis varies according to these groupings. Cancer 2004. © 2004 American Cancer Society.
Follicular thyroid carcinoma (FTC) accounts for approximately 10% of all thyroid malignancies.1 The term follicular carcinoma was used initially because of the microscopic follicular pattern, consisting of small follicles without papillae, exhibited by this malignancy. FTC is more common in areas of endemic goiter.2 It reportedly is more prevalent in African Americans than in Asians or Caucasians.2 The female-to-male ratio for individuals with FTC is 3.3:1.0, and FTC occurs most often in patients between ages 45–49 years and 60–70 years.1, 2 FTC appears to be decreasing in incidence, at least in the United States.3 This decrease probably is due to more accurate histologic diagnoses; that is, the identification of mixed papillary thyroid carcinoma (PTC) and follicular variants of PTC as PTC rather than FTC; and also because of the eradication of iodine deficiency.3 Some experts suggest that FTC arises from follicular adenomas.4
FTC generally is considered to be a more aggressive tumor than PTC, because patients with FTC often present with higher-stage disease,5 but patients with FTC and PTC have similar prognoses when they are matched for age and disease stage.6, 7 Lymph node metastases are uncommon in patients with FTC, occurring in ∼10%, whereas ∼50% of patients with PTC have at least microscopic lymph node metastases. Distant pulmonary or bone metastases, however, are more common in patients who have FTC (30%) compared with patients who have PTC (15%).8, 9
Hürthle cell carcinomas (HCCs) are considered follicular tumors according to the World Health Organization classification system,10 although others,11, 12 including our group,13 propose that HCCs represent a distinct entity with different behavior. Therefore, we excluded HCCs from the current study. HCC occurs more often in iodine-rich areas, whereas FTC occurs more often in iodine-poor areas. HCC occurs more often after exposure to therapeutic low-dose radiation compared with FTC and is more likely to be familial. HCC also is more likely to involve regional lymph nodes than FTC.13 Only ≈ 7% of HCCs take up radioiodine, compared with ∼75% of FTCs, although both typically produce thyroglobulin.
In 1914, Graham14 was the first to provide a histologic definition of FTC based on the criteria of vascular and capsular penetration. However, considerable differences of opinion remain among endocrinologists, pathologists, and endocrine surgeons about the precise criteria that should be used to diagnose minimally invasive and widely invasive FTC.15 Some experts include patients who have follicular tumors with angioinvasion in the low-risk group, and others include such patients in the high-risk group. This situation makes it difficult to compare reports from different medical centers regarding tumor characteristics and prognoses.
In 1993, we reported our experience with 65 patients who were treated at the University of California–San Francisco Medical Center between 1956 and 1990.9 We wanted to update our experience up to August 2000, because as of that date, we had cared for 132 patients. The objective of the current investigation was to determine whether there is a direct relation between the histologic degree of invasion of FTC and tumor recurrence and patient survival. We propose that minimally invasive FTC should be categorized as FTC with capsular invasion only, moderately invasive FTC should include angioinvasion with or without capsular invasion, and widely invasive FTC should include tumors that invade directly into adjacent muscle and/or parenchyma. We believe that for a histologic classification scheme to be useful, it should be correlated with tumor behavior. We also suggest that there should be a generally accepted histologic classification for FTC, so that results regarding survival and disease stage, as reported from different medical centers, can be compared.
MATERIALS AND METHODS
One hundred thirty-two patients with FTC underwent a combined total of 182 thyroid carcinoma–related operations. Patients were followed for a mean of 7.9 years and a median of 6 years (range, 2 months to 39 years). There were 92 women (70%) and 40 men (30%), with a female-to-male ratio of 2.3:1. The mean age at presentation was 49 years (range, 9–90 years), and incidence peaked at 46 years. Twelve patients (9%) had a history of radiation exposure, and no patient had a family history of thyroid carcinoma.
Most patients presented with an asymptomatic, solitary thyroid nodule, 25 patients (19%) had a dominant nodule in a multinodular goiter, 4 patients had symptoms of hoarseness, 3 patients had dysphagia, and 4 patients had pressure in the neck. Three patients exhibited hypothyroidism, and all others had normal thyroid activity. Twelve patients had documented distant metastases prior to their initial thyroid operation.
The operative treatments for these patients are listed in Table 1. Sixty-six patients underwent an initial lobectomy, usually after a fine-needle aspiration biopsy with cytology that revealed a follicular neoplasm. Forty-four of those patients underwent a completion total thyroidectomy within 6 months of their initial thyroid lobectomy. Forty-five other patients underwent an initial total thyroidectomy, 12 of those patients underwent an ipsilateral modified radical neck dissection with or without mediastinal lymph node dissection, 3 other patients underwent an upper mediastinal lymph node dissection, and 1 patient underwent a bilateral modified neck dissection. Seventeen patients underwent an initial subtotal thyroidectomy, and 4 of those patients subsequently underwent completion total thyroidectomy. One patient underwent lumpectomy and subsequently underwent completion total thyroidectomy. Two patients underwent surgical resection of local metastases (one patient had central neck metastases, and one had lateral neck metastases), and a third patient underwent resection of a distant metastasis (to bone).
|Initial surgery||No. of patients (%)||No. completed|
|Total thyroidectomya||45 (34.0)||—|
|Subtotal thyroidectomy||17 (13.0)||4|
For the current investigation, we used the following classification to describe the invasiveness of FTC: 1) minimally invasive FTC was characterized by the unequivocal observation of tumor cell invasion through the entire thickness of the tumor capsule into either benign thyroid tissue or other tissue (tumors with apparent invasion into but not through the capsule were considered follicular adenomas); 2) moderately invasive FTC was characterized by the observation of thyroid carcinoma cells in a vein or veins either in or immediately adjacent to the tumor capsule, with or without capsular invasion; and 3) widely invasive FTC was characterized by the observation of a broad area or areas of transcapsular invasion of tumor into the extrathyroidal tissue. The classification was based on review of the final pathology reports, although many of the slides were reviewed by a single pathologist (P.T.).
Complications occurred in 9 of the thyroid operations (5%). Two patients developed hypoparathyroidism, although one of those patients had this complication before reoperation at our medical center. Two patients had recurrent laryngeal nerve dysfunction, one of whom purposely had the nerve resected due to tumor invasion, and one patient also had an external laryngeal nerve injury. Two patients developed postoperative hematomas, one with a keloid and one with a wound breakdown. Six of these complications occurred at the time of the first operation, and three occurred at the time of reoperation.
Sixty-four patients received postoperative iodine-131 (131I) ablative therapy; 8 patients received a second dose, and 2 patients received a third dose. The individual doses of 131I ranged from 50 millicuries (mCi) to 200 mCi. Six patients received external radiotherapy as well as 131I therapy directed against lung, bone, liver, and brain metastases. Two other patients underwent resection of their metastatic tumors and received external radiotherapy, and two patients received both external radiotherapy and chemotherapy. One other patient received chemotherapy, and four patients underwent surgical resection of distant metastases without radiation or chemotherapy. Follow-up data regarding recurrences and survival were obtained from personal physician records as well as from the University of California–San Francisco Cancer Registry and the California Tumor Registry. The Tumor Registry contacts patients and primary physicians annually to ascertain disease status. The current investigation was approved by the University of California Committee on Human Research.
Twelve of 132 patients (9%) presented with distant metastases, 8 patients (6%) presented with lymph node metastases, and 112 patients (84%) presented with tumors that were clinically confined to the thyroid gland. Distant metastases were present in the bone (5 patients), lungs (4 patients), pituitary gland (1 patient), ovary (1 patient), and scalp (1 patient).
Pathology reports or slides were available for 119 of 132 patients (90%) in the series, and our descriptions of minimally invasive FTC, FTC with capsular invasion, and widely invasive FTC focused on this subgroup. Forty-five of 119 patients (38%), presented with minimally invasive disease and capsular invasion only, 10 patients (8.4%) presented with angioinvasion only, and 40 patients (33.6%) presented with both capsular invasion and angioinvasion. Twenty-four patients (20%) presented with widely invasive FTC.
Among patients with distant metastases at presentation, one patient had minimally invasive FTC with capsular invasion only, four patients had moderately invasive FTC with angioinvasion and capsular invasion, and seven patients had widely invasive FTC. Among patients with lymph node metastases, one patient had minimally invasive FTC, and seven patients had widely invasive FTC. No patients with moderately invasive FTC had lymph node involvement.
Nine patients from the subgroup of 119 had coexisting thyroid carcinoma: 8 patients had distinct PTC, separate from the FTC, and 1 patient had a coexisting medullary thyroid carcinoma (MTC). In 4 patients (3%) with FTC, another focus of FTC was found in the contralateral lobe at the time of completion total thyroidectomy. In 2 of those 4 patients, the contralateral FTC was among multiple tumor nodules, while in another patient, the contralateral FTC measured 3.5 cm. The patient with FTC and MTC died 20 years after initial treatment from metastatic FTC to bone without recurrent MTC.
When the 12 patients who presented initially with distant metastases were excluded, 21 patients (16%) from the entire group of 132 patients developed recurrent metastases 6 months after their initial treatment. These included local recurrence in seven patients, distant recurrence in seven patients, and local and distant recurrence in seven other patients.
For the subgroup of patients with complete pathology data (n = 119), a Kaplan–Meier survival curve is illustrated in Figure 1. Among the 45 patients who had minimally invasive FTC and capsular invasion only, 6 patients (13%) developed recurrent or persist disease, and 5 patients (11%) died. Among the 50 patients who had angioinvasion, 40 patients also had capsular invasion, and 10 patients did not. Ten patients (20%) developed recurrent or persistent disease, and 7 patients (14%) died. Although the differences in survival were small between the two groups, patients with angioinvasion, with or without capsular invasion, had a poorer prognosis (P < 0.0001). Among the 24 patients who had widely invasive disease, 9 patients (38%) developed recurrent disease, and 8 patients (33%) died. All 7 patients with widely invasive disease and persistent disease died, for an overall death rate at 5 years of 62% in this group.
When the 12 patients who initially presented with distant metastases were excluded, only 4 deaths (8.8%) were observed at follow-up among patients with minimally invasive FTC and capsular invasion (see Table 2). Of the 50 patients with moderately invasive FTC who initially presented without distant metastases, 5 (10%) died. Surprisingly, the presence of persistent disease at presentation did not appear to have an adverse influence on survival in patients with either minimally invasive or moderately invasive FTC. With the data plotted for cumulative hazard, the relative risk among these three groups can be seen in Figure 1.
|Tumor invasion||No. of patients (%)|
|Recurrent disease||Persistent disease||Dead without preoperative metsa||Total deaths due to FTC|
|Capsular invasion (45 patients)||5 (11)||1 (2.2)||4 (8.8)||5 (11)|
|Angioinvasion with and/or without capsular invasion (50 patients)||6 (12)||4 (8)||5 (10)||7 (14)|
|Wide spread invasion (24 patients)||9 (37.5)||8 (33.3)||8(32.3)||15 (62.5)|
Overall, 11 of 12 patients (92%) who had persistent disease and 17 of 21 patients (81%) who developed recurrent FTC died (Fig. 2). Although, unfortunately, survival was poor in both patients with persistent FTC and patients with recurrent FTC, patients who had persistent disease had a poorer prognosis compared with patients who had recurrent FTC (P < 0.0001).
Among the entire group of 132 patients, women had a slightly better prognosis compared with men (P < 0.01) (Fig. 3). Causes of death represented in this series include 28 patients (21%) who died of FTC, 14 patients who died of unrelated causes, and 6 patients who died of other malignancies. For 5 patients, the cause of death was unknown (P < 0.005).
Overall, 12 of 45 patients (27%) who had minimally invasive FTC, 31 of 50 patients (62%) who had moderately invasive FTC, and 24 of 24 patients (100%) who had widely invasive FTC underwent total thyroidectomy (n = 45) or completion total thyroidectomy (n = 50). Sixty-four of those patients received a treatment dose of 131I, and some patients received several doses. Although the objective of our investigation was not to analyze the impact of total thyroidectomy and radioiodine ablation on survival, we would advocate this treatment for all patients who have increased thyroglobulin levels after undergoing thyroidectomy and for patients who have moderately and widely invasive FTC.
Table 3 lists the reports in the literature regarding the outcome of patients with minimally invasive FTC. Some authors15–17 included only tumors with capsular invasion as minimally invasive FTCs, whereas others included patients with angioinvasion with or without capsular invasion as minimally invasive FTCs.9, 18–25 Others suggest that FTCs with angioinvasion behave in a more aggressive manner than FTCs with capsular invasion only.26, 27 The recurrence rate in these series ranged from 5% to 42.8% (mean, 18.2%), and the death rate from thyroid carcinoma ranged from 2% to 42.8% (mean, 13.9%). Four of 57 patients (7%) who had capsular invasion developed recurrent disease, compared with 57 of 399 patients (17%) who had angioinvasive FTC with or without capsular invasion (P = 0.07). In patients who had widely invasive FTC (Table 4), the recurrence rate ranged from 33% to 90% (mean, 55.8%), and the death rate ranged from 16% to 81% (mean, 50.2%).
|Author(s)||Follow-up period||No. of patients||Type of resection||Pathology||No. of patients (%)|
|Warren, 193118||1923–1927||34||NA||Angioinvasion||NA||2 (5.8)|
|Lindsay, 196019||1920–1954||32||NA||Capsular, angioinvasion||NA||3 (9.0)|
|Khan and Perzin, 198320||1915–1972||35||Unilateral-bilateral lobar||Capsular, angioinvasion||9 (26)||NA|
|Evans, 198415||10 yrs||7||Unilateral-bilateral lobar||Capsular||3 (42.8)||3 (42.8)|
|Crile et al., 198521||1948–1977||48||Unilateral-bilateral lobar||Capsular, angioinvasion||NA||NA|
|Cady et al., 198516||1961–1980||30||Unilateral lobar||Capsular||0 (0)||0 (0)|
|Lang et al., 198622||1969–1980||106||NA||Capsular, angioinvasion||8 (7.5)||NA|
|van Heerden et al., 199217||1971–1985||20||Unilateral-bilateral lobar||Capsular||1 (5.0)||NA|
|Emerick et al., 19939||1956–1990||48||Unilateral-bilateral lobar||Capsular, angioinvasion with/without capsular||3 (6.2)||1 (2.0)|
|DeGroot et al., 19955||1968–1993||36||Unilateral-bilateral lobar||NA||9 (25.0)||5 (13.8)|
|Schmid et al., 199723||1970–1987||119||Unilateral-bilateral lobar||Capsular, angioinvasion with/without capsular||16 (13.5)||7 (5.9)|
|Besic et al., 199924||1972–1992||72||Bilateral lobar||Capsular, angioinvasion||NA||NA|
|Zidan et al., 200025||1969–1992||36||Bilateral lobar||Capsular, angioinvasion||NA||NA|
|Author(s)||Follow-up period||No. of patients||No. of patients (%)|
|Lindsay, 196019||1920–1954||25||15 (60)||17 (68)|
|Khan and Perzin, 198320||1915–1972||47||17 (36)||17 (36)|
|Evans, 198415||10 yrs||11||10 (90)||9 (81)|
|Crile et al., 198521||1948–1977||16||NA||NA|
|Lang et al., 198622||1969–1980||61||20 (33)||10 (16)|
|DeGroot et al., 19955||1968–1993||5||3 (60)||—|
|Schmid et al., 199723||1970–1987||200||NA||NA|
The current retrospective investigation and review of the literature demonstrated that patients with minimally invasive FTC (capsular invasion only) had slightly but significantly better survival (97.8% at 10 years) compared with patients who had angioinvasion with or without capsular invasion (80% at 10 years). Both groups had a more favorable prognosis compared with patients who had widely invasive FTC (37.5% at 10 years; P < 0.0001) (Fig. 1).
Patients who developed recurrent FTCs often appeared to be tumor free for at least 6 months and had a more favorable prognosis compared with patients who presented with distant disease or who had documented FTC within 6 months of initial surgical treatment (Fig. 2). Survival also was better in women compared with men (Fig. 3). In the current retrospective review, we did not determine whether the 95 patients who underwent total thyroidectomy and the 66 patients who also received ≥ 1 dose of radioiodine therapy had fewer recurrences and better survival. Because most of these patients had moderately or widely invasive disease, rather than minimally invasive disease, it is difficult to determine the impact of treatment, because the number of patients for comparison is relatively small.
Our findings, which are supported by some reports in the literature, suggest that patients with FTC should be separated into three groups based on histologic disease status at diagnosis—minimally invasive (capsular only), moderately invasive (angioinvasion with or without capsular invasion), and widely invasive—because these groupings are correlated with patient survival. Unfortunately, our data also suggest that some patients with only capsular invasion die of FTC. This finding is in disagreement with the findings of van Heerden et al.17 but is consistent with results obtained by others.9, 15, 23 Considerable differences of opinion exist in the literature regarding the precise definitions of FTC and minimally invasive FTC.17, 20, 23, 28 Patients with follicular neoplasms that do not exhibit invasion extending completely through the capsule are considered to have benign tumors. Patients with such neoplasms rarely develop or have distant metastases.11, 12 Invasion through the capsule, therefore, is necessary for a follicular neoplasm to be considered a minimally invasive FTC.11, 12
Although it appears to be logical to separate FTC into three risk groups based on histologic criteria, some experts have not accepted this recommendation.29 In the current investigation, we did not address other factors that influence survival, such as type of treatment, patient age, and gender, and we did not evaluate the importance of TNM status or other classifications in predicting tumor behavior.
In conclusion, we recommend that FTCs be classified as minimally invasive (capsular invasion only), moderately invasive (angioinvasion with or without capsular invasion), or widely invasive, because this histologic classification system has been found to be correlated with disease recurrence and survival in patients in the current study and in patients discussed in the literature. We also recommend the use of uniform histologic criteria for diagnosing FTCs, so that studies conducted at different medical centers can be compared.
The authors thank Dr. Gennaro D'Auria, general manager at ASL-NA3, Ospedale San Giovanni di Dio, Frattamaggiore, Italy, for his mentorship of and financial assistance to Dr. D'Avanzo. The authors also thank the University of California–San Francisco Cancer Registry and the California Tumor Registry for their support in conducting patient follow-up; Ms. Kate Poole for her excellent clerical assistance; and the Friends of Endocrine Surgery, the Gerald Heller Family Foundation, the Albert Clark Family Foundation, the Sanford and Helen Diller Foundation, and Mount Zion Health Systems for their financial support.
- 10Histological typing of thyroid tumors. In: HedingerCE, editor. International histological classification of tumors. Volume 11. Berlin: Springer-Verlag, 1988., , .
- 11Pathobiology of the parathyroid and thyroid gland. Baltimore: Lippincott Williams & Wilkins, 1993..
- 12Tumors of the thyroid gland. Fascicle 5. Third series. In: RosaiJ, CarcangiuM, De LellisRA. Atlas of tumor pathology. Washington, DC: Armed Forces Institute of Pathology, 1992: 161–182., , .
- 13Hurthle cell carcinoma. Cancer Control. 1997; 4: 13–17., .
- 14Malignant epithelial tumors of the thyroid with special reference to invasion of blood vessels. Surg Gynecol Obstet. 1924; 42: 781–790..
- 18The significance of invasion of blood vessels in adenomas of thyroid gland. Arch Pathol. 1931; 11: 255–257..
- 19Carcinoma of the thyroid gland. Springfield, IL: Charles C. Thomas, 1960..
- 26Histopathologic prognostic factors in follicular thyroid carcinoma: an attempt to quantify grading. Mod Pathol. 2000; 13: 74A., , , , , .
- 28Atlas of tumor pathology. Tumors of the thyroid gland. Washington, DC: Armed Forces Institute of Pathology, 1968: 84–95., .