Keywords

Introduction

The Hürthle cell variant of follicular carcinoma is composed of large acidophilic, Askanazy, or oncocytic cells that are considered altered follicular cells [1–10]. For instance, Hürthle cells bind thyrotropin (TSH) and have TSH receptors, like other types of follicular thyroid cells [11, 12]. Similar to other follicular neoplasms, Hürthle cell carcinoma is more common in women than men, but the patients tend to be older than those with typical follicular thyroid carcinoma (FTC; see Chap. 70). Hürthle cells contain many mitochondria (which are the basis for the abundant, eosinophilic, and granular cytoplasm), frequently eccentric nuclei, and visible nucleoli. The Hürthle cell carcinoma variant of follicular carcinoma discussed here is to be distinguished from the variant of papillary thyroid cancer, “Hürthle cell papillary thyroid carcinoma,” which also contains an abundance of oxyphilic cells [13]. The central genetic or environmental factors that allow a thyrocyte to differentiate into a Hürthle cell are unknown, and Hürthle cells may occur in a variety of thyroid disorders. Solitary thyroid nodules may have a predominance of Hürthle cells, to the exclusion of more typical thyrocytes, and these lesions are often highly cellular with minimal colloid. When such a lesion is aspirated for cytologic examination, the interpretation would likely be “suggestive of a Hürthle cell neoplasm,” because as with follicular tumors, a definite diagnosis of benign or malignant would be difficult or impossible to make. Hürthle cell neoplasms may be benign or malignant, and the distinction is based on the demonstration of vascular or capsular invasion, metastatic capacity, and growth rate, similarly to other follicular neoplasms [7, 14, 15].

When patients with Hürthle cell carcinoma are stratified regarding low vs high risk, there does not appear to be a significantly worse prognosis than that for follicular carcinoma; see Chaps. 70 and 79 [16, 17]. In contrast to other follicular carcinomas, they have a higher rate of bilaterality or multi-centricity.

With the use of fine-needle aspiration (FNA) cytology, some solitary thyroid nodules may also have a more varied appearance, where Hürthle cells intermingle with thyrocytes, macrophages, and lymphocytes and have moderate amounts of colloid. In such circumstances, the cytology may be more difficult to interpret, and when insufficient Hürthle cells are present, the tumor is characterized as a Hürthle cell neoplasm. Furthermore, Hürthle cells may be found in the thyroid glands of patients with Hashimoto’s thyroiditis and other benign thyroid disorders but usually in this circumstance, the Hürthle cells are scattered and are not the predominant cell type [7, 18]. In one study, age greater than 65 years and an elevated preoperative serum thyroglobulin level aided in the prediction of a follicular neoplasm being malignant [8].

The typical Hürthle cell neoplasm is composed mostly of these distinctive cells and is generally considered to be a variant of follicular carcinoma. Recognizing the malignant potential of a tumor depends on the evidence of aggressive behavior at its periphery [19–23]. Bizarre, large, and/or hyperchromatic nuclei may be a striking histological feature, and these are more common than in benign proliferations of oxyphilic cells.

Metastases to cervical lymph nodes are more frequent than with the usual follicular carcinoma, especially after the patient has undergone surgery. Notwithstanding the comments above, there are some studies that suggest oxyphilic follicular carcinomas are more aggressive than typical nonoxyphilic follicular carcinomas. Particularly, the presence of nondiploid cells in an oxyphilic carcinoma indicates a poorer prognosis than that with diploid nuclei [24]. Papillary carcinomas may also contain significant populations of these cells [13]. Whether they are more aggressive than a nonoxyphilic cancer with otherwise similar characteristics remains uncertain.

However, what does appear clear is that without specific attention to risk stratification, Hürthle cell malignancies will tend to have a worse prognosis than other follicular tumors in many [24–26], but not all, studies [27]. This may be partly because of their greater tendency to be locally invasive and propensity to concentrate radioiodine less avidly, thereby rendering them more difficult to manage with isotopic scanning and therapy [28, 29]. Yet, this is not universal, and some Hürthle cell cancers will trap iodine [30]). In a review of 89 cases of Hürthle cell carcinoma seen at the MD Anderson Cancer Center, Lopez-Penabad et al. [31] expressed a rather pessimistic assessment of the efficacy of currently available therapies for this variant of FTC. There was a 40 % cause-specific mortality over their long follow-up interval, with no improvement in mortality rates found over the past 50 years. Larger tumor size and more advanced age were negative prognostic indicators. Additionally, in the setting of metastatic disease, there was no further survival advantage seen with more extensive surgery, external radiation therapy, chemotherapy, or even radioactive iodine therapy. Mills et al. noted that the stage of disease, the existence of metastatic lesions, and the presence of cervical lymph node disease affected outcome in patients with Hürthle cell carcinoma [32].

Clinical Presentation

Hürthle cell carcinomas may represent about 3–5 % of all types of thyroid carcinomas. Most Hürthle cell carcinomas appear to be a more aggressive kind of follicular carcinoma, with more frequent recurrences, higher morbidity, and higher mortality [7, 14, 18, 24, 26, 31]. The tumors are frequently multifocal and bilateral.

Thompson and associates [14, 19, 20] suggested that it is difficult to differentiate benign from malignant Hürthle cell tumors. The implication of their studies is that even experienced pathologists may not be able to make a reliable distinction. Carcangiu and coworkers [33] and Grant and associates [22] support the concept that strict histological criteria and adequate sampling may be able to differentiate Hürthle cell carcinoma from adenoma in nearly all cases. Grant et al. [22] reviewed the world literature and observed that only 6 of 642 patients with apparent benign Hürthle cell adenomas were found to have a recurrence, thereby indicating that the tumor was a carcinoma, with an incidence of less than 1 %. Gosain and Clark [23] found no patients with Hürthle cell adenoma in whom recurrences were observed. Similarly, Bondeson and coworkers [24] studied 42 patients diagnosed with Hürthle cell adenoma over a 2–20-year period and found no recurrences.

The above reports help to support the contention that Hürthle cell adenomas can be accurately diagnosed and distinguished from carcinoma on histological examination by experienced pathologists. As with other follicular carcinomas, the major histological criteria that separate a Hürthle cell adenoma from a carcinoma are vascular and/or capsular invasion. However, subtleties do remain. For example, does the capsular invasion have to be completely through the capsule, or is invasion into, but not through, the capsule enough to make the diagnosis? It is also important to ensure that sufficient histologic sections were taken and examined. Notwithstanding these concerns, it is likely that experienced pathologists can reliably make this differential diagnosis.

The absence of radioiodine uptake by residual or metastatic follicular cancer renders management much more difficult, but not impossible. Typically, these tumors are less avid for radioactive iodine and therefore respond less often than the usual follicular carcinoma. Several groups are currently evaluating the thyroidal sodium iodide symporter (Na+/I− symporter function and expression in Hürthle cell carcinomas relative to FTC [11, 12, 18]. These studies are exploring methods for the management of follicular cancers that have lost their ability to either trap iodide (and be treatable with radioiodine) or to synthesize and release thyroglobulin. The ability of both normal and malignant thyroid cells to concentrate iodide is dependent on expression of the NIS gene [34–36]. The loss of this gene during tumor dedifferentiation can account for the tumor’s failure to concentrate iodide. Gene therapy or redifferentiation therapy with retinoic acid, depsipeptide, and other agents had been thought to hold promise for the restoration of both radioiodine uptake for potential treatment and thyroglobulin production for monitoring recurrence [37–41]. However, further analysis of results with retinoic acid was less than optimistic [42].

It is reasonable to assume that one major reason why these tumors do not respond as well to therapy is because they do not concentrate radioiodine as well as usual follicular carcinomas. It may be appropriate to approach Hürthle cell carcinomas as if they were medullary carcinomas, i.e., with more aggressive diagnostic procedures and treatment [23, 43].

In most clinical circumstances, patients diagnosed with a Hürthle cell neoplasm by FNA should undergo surgery somewhat promptly. We recommend a near-total to total thyroidectomy by an experienced thyroid surgeon. It is important to discuss with the patient the alternative approaches of a near-total thyroidectomy compared to a lobectomy with isthmusectomy [44]. If only a lobectomy and isthmusectomy are performed, and if the lesion is found to be cancerous, then a subsequent completion thyroidectomy must be performed. This completion thyroidectomy frequently causes mental and psychosocial distress to the patient, especially if the requirement for this procedure is not expected by the patient. However, only about 20 % of Hürthle cell neoplasms diagnosed by FNA are found to be malignant. If a total thyroidectomy is conducted initially, then in 80 % of cases, this procedure would be unnecessarily aggressive and exposes the patient to a higher risk of temporary and permanent hypocalcemia, along with recurrent laryngeal nerve paralysis.

The decision regarding which operation is needed for a patient with a solitary thyroid nodule and an FNA consistent with Hürthle cell neoplasm is difficult. A frozen-section interpretation is often problematic and therefore not helpful [45]. It is important to candidly discuss the advantages and disadvantages of each approach with the patient and family and arrive at a mutual decision. The initial operation should include an ipsilateral central node dissection. Obviously, the surgeon must be allowed to exercise judgment at the time of surgery about the precise operative procedures. It has been suggested [11] that a routine modified radical neck dissection be performed when the tumor is found in the central compartment or cervical nodes or, alternatively, a total thyroidectomy [44].

Because these tumors are somewhat less differentiated than most papillary thyroid cancers and FTCs, there is a reasonable chance that they can be detected by 18-fluorodeoxyglucose positron emission tomography (FDG-PET) when they do not concentrate radioiodine. Lowe et al. [46] found PET scanning to be useful to identify both local and metastatic disease, thereby facilitating disease management. Growth of Hürthle cell tumors has been described to reflect the net of proliferative vs apoptotic indices, and it may become feasible to exploit these characteristics to distinguish benign from malignant lesions [47].

McDonald and coworkers [48] reviewed 40 cases of Hürthle cell carcinoma, noting that this number represented 4 % of all thyroid cancers in their experience. Their median follow-up interval after thyroidectomy was 8.5 years. Vascular or capsular invasion was observed in 32 patients, extrathyroidal invasion in 11, and regional lymph node involvement in 2. One patient had distant metastases at presentation, and only nine patients received 131I. Of 34 subjects analyzed, 5 died of thyroid cancer, 9 died of nonthyroidal causes, 4 were alive with existing disease, and 16 were alive without evidence of disease. At about an average of 4 years, nine patients had recurrences and five had distant disease. Recurrent disease was associated with mortality in half of these patients. Risk factors assessed at initial presentation were useful to help predict recurrence. Low-risk tumors did not recur (e.g., tumors less than 5-cm diameter, lack of distant metastases, men younger than 41 year, and women younger than 51 year). Tumor size and the presence of distant metastases were more important prognostic indicators than age in one recent series of patients [49].

Bhattacharyya [50] performed a retrospective review of the Surveillance, Epidemiology, and End Results (SEER) database for cases between 1973 and 1998, finding that 3 % of cases represented Hürthle cell carcinoma, and 555 of the patients (377 women, 178 men) were analyzed. Outcomes in 411 were compared to outcomes in 411 matched patients with follicular carcinoma; 5- and 10-year mortality rates were 15 % and 29 % vs 11 % and 45 %, respectively. The survival time was also not different, with an average of 109 months for Hürthle cell carcinoma and 113 months for follicular cancer. For the patients with Hürthle cell carcinoma, increased mortality was associated with larger tumor size and male gender, but not the presence of local invasion. In a small patient series, Lopez-Penabad et al. [31] observed the worst prognosis in older patients with larger tumors and local extension. Similar findings were reported by Kushchayeva et al. in a series of 33 patients [51].

Because some Hürthle cell carcinomas were found to have ret/PTC gene rearrangements similar to papillary thyroid cancers [52], as well as a propensity to spread to local lymph nodes like papillary cancers, it may be that there can be subspecies of what we have presumed to be classic Hürthle cell carcinoma that represent variants of either follicular or papillary thyroid cancer [13]. Such differences might account for our difficulty in comparing individual reports in the literature with mortality and morbidity rates, of which some but not all support the view that Hürthle cell carcinoma is linked with a poorer prognosis than the usual FTC or papillary thyroid cancer.

Following appropriate surgery for Hürthle cell carcinoma , radioiodine scanning and therapy is recommended. Scan preparation would be routine and is usually performed about 6 weeks after surgery. Preparation by the use of levothyroxine withdrawal or stimulation by the use of rhTSH can be utilized. A diagnostic radioiodine scan is important before therapy to help determine the avidity of the remaining thyroid cells for radioiodine and to define the nature and extent of remaining thyroid tissue or disease. Assuming that there is visible uptake, the diagnostic scan is then followed by radioiodine therapy, usually with 100–150 mCi 131I. A posttreatment scan is performed approx 7–10 day after treatment. (These protocols are detailed in Chaps. 11, 19, 33 and 34) rhTSH stimulation is not presently approved by the FDA for use in patients with metastatic thyroid cancer, although two recent studies have suggested it may be as effective as levothyroxine withdrawal [53, 54].

As described above, many Hürthle cell cancers will not trap radioiodine, and sometimes only as little as 10 % of Hürthle cell cancers will trap and respond to radioiodine. This number seems low in our experience and, of course, somewhat depends on the dose of 131 I used for scanning, the length of time that the patient did not receive thyroid hormone, the extent of TSH elevation, and possibly the assiduous adherence to a low-iodine diet. Based on published reports, it may be difficult to adequately assess these factors. Perhaps in some cases, lack of apparent iodine avidity by the tumor may not be an accurate representation of the tumor’s true properties.

For surveillance over the subsequent 5 years after initial surgery and ablation, we recommend following the patient with physical examinations, thyroid function tests, and thyroglobulin monitoring about every 3–6 months for the first several years, and possibly every 4–6 months for the next several years if there has been no evidence of disease recurrence. Our approach of utilizing radioiodine scans is consistent with that of Besic et al. [30]. In the initial year or two of surveillance, levothyroxine therapy is used in a dosage designed to achieve suppressed TSH in most patients (0.1 μU/mL or lower), depending on the clinical context. Thyroglobulin levels are analyzed at the same time as thyroid function tests, and the thyroglobulin level during L-thyroxine suppression must be less than 2 ng/mL (according to the assay). rhTSH stimulation testing may be employed as outlined by a consensus group of thyroid cancer investigators [55] The latter follow-up would generally include a repeat whole-body 131I scan in 1 year, then again 3–5 years later. Given the aggressive nature of this tumor, we may also obtain occasional radiographs of the chest. An imaging study of the neck, such as a magnetic resonance image or sonogram, is routinely performed, especially if the tumor is not iodine-avid, if the thyroglobulin level is increasing, or if palpable cervical abnormalities become manifest. In women, particularly those who are postmenopausal, suppressive doses of levothyroxine therapy should be accompanied by measures to prevent osteoporosis (daily oral ingestion of 1–1.5 g of calcium, 400 U of vitamin D, exercise against gravity, and the possible addition of a bisphosphonate as appropriate.

Other scanning agents may have reasonably good utility for the detection of Hürthle cell carcinoma. In a study comparing radioiodine and thallium scanning to that with 99mTc-MIBI, Yen et al. [56] reported a 100 % specificity and an 82 % sensitivity for 99mTc-MIBI in patients with Hürthle cell carcinoma. The utility of FDG-PET scanning has already been mentioned [46] and was also reported earlier [57] to have an 80 % specificity and a 92 % sensitivity for this tumor. The issue of which scanning agent to employ arises when serum thyroglobulin levels indicate residual or recurrent disease. Despite the proven value of PET scanning, because of its cost and lack of widespread availability, we believe that radioiodine scanning should be attempted first, followed by 99mTc-MIBI, before utilizing FDG-PET.

Hürthle cell cancer is discussed further in Chaps. 55, 56, and 57. These and other follicular cancers that do not concentrate radioiodine may be treated with chemotherapy (see Chaps. 64 and 99), external radiation therapy [58–61]; see Chaps. 63, 78, 88 and 101, or redifferentiation therapy with retinoic acid or other agents could be attempted [37–39, 62]. However, although external radiation therapy may cause apparent tumor regression and provide palliation and reduced recurrence rate, even in Hürthle cell carcinoma [63], the effect may be transitory with little improvement in survival rate [64].

Recent advances in molecular analysis of thyroid FNA samples and thyroid histopathology samples hold promise of improved ability to predict the presence of thyroid cancer as well as its potential aggressiveness [3, 65–67].