Nature of the Disease

Anaplastic thyroid carcinoma (ATC) is the most aggressive and deadly thyroid malignancy, as well as one of the deadliest solid tumors in humans. Although it is rare, constituting less than 2% of incident thyroid carcinomas, it is likely increasing in incidence, coincident with the dramatic increase seen in thyroid malignancies (mostly papillary carcinomas) since the 1990s. One study suggests that ATC may be diminishing as a consequence of better treatment of preceding differentiated thyroid cancers¹ and in some regions due to increased dietary iodine;² however, it is our impression that the number of ATC cases in North America has escalated. This rapidly progressive cancer is resistant to all known cytotoxic agents and represents the epitome of lethality in solid tumors, mandating an equally aggressive clinical approach.

The best of evidence affirms that ATC results from a terminal dedifferentiation of preexisting carcinomas of the thyroid follicular cell (see Chapter 17, Molecular Pathogenesis of Thyroid Neoplasia). Whereas the consistent observation of coexisting follicular or papillary thyroid cancers with ATC has long suggested this association,^3,4 modern genetic analysis has further substantiated this hypothesis, showing shared genetic alterations in the differentiated thyroid cancers and the ATCs found in the same patients.^5,6 One fairly consistent biochemical feature distinguishing ATC is the enhanced expression of p53-induced nuclear protein 1⁷; another is a specific pattern of down-regulated micro-RNAs,⁸ and further findings include expression of a novel gene, called overexpressed in anaplastic thyroid carcinoma-1 (OEATC-1),⁹ as well as overexpression of UbcH10,¹⁰ among others. An excellent review by Smallridge et al.¹¹ summarized a wide range of such discoveries. Nonetheless, the likely pathogenesis of ATC from differentiated thyroid cancers places a particular burden on clinicians to assertively diagnose and treat differentiated thyroid cancers before they become lethal ATC.

ATC is typically rapidly progressive and we have observed tumor doubling times as brief as 3 days. This disease is not amenable to typical clinic scheduled openings and physician vacations, warranting a “stat” approach on the part of the involved physicians. Although there have been many research innovations and trials of promising treatments, the clinical outcomes for optimally treated ATC patients have not significantly changed since we last summarized this topic in the late 1990s.¹² Median survival of patients receiving diverse therapeutic interventions ranged from 3 to 5 months in recent series,^13–15 although aggressive multimodal therapy significantly enhances survival when the primary tumor is resectable.¹⁶ A review of 516 ATC patients reported in the National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) database between 1973 and 2000 revealed the cause-specific mortality to be 68% at 6 months and 81% at 1 year.¹⁷ ATC progresses inexorably, causing rapid death if treatment cannot gain local control and delayed mortality if excellent local control permits distant disease to gain the upper hand. The rare ATC patients who survive more than 2 years are generally considered cured of their ATC; therapeutic attention should then be directed toward their underlying differentiated thyroid cancers. On the other hand, our clinical experiences suggest that a skilled “therapeutic team” can improve this dismal situation by a timely and assertive application of multimodal therapy.

Presentation and Pathology

ATC is most commonly found in the context of a rapidly growing thyroid mass, often with symptoms of hoarseness, respiratory obstruction, dysphagia, and neck pain.^18–20 Primary tumor sizes range from a mean of 5 to 8 centimeters in diameter.^15,17,21,22 More than a third of patients present with a long-standing goiter (presumably consequent to an undiagnosed differentiated cancer) that suddenly enlarges.³ Alternatively, a patient under therapy for papillary or follicular thyroid cancer will reveal sudden progression of metastatic sites that reveal ATC upon biopsy or resection. Although the median age at presentation is in the seventh decade,^13–15,23 there are rare patients as young as 14 or 26 years of age.^24,25 More than 40% of patients have cervical lymphadenopathy and 30% have spontaneous vocal cord paralysis.^3,19,21 The two- to threefold predominance of female patients over men^18,23 likely reflects the female predominance of the underlying differentiated progenitor thyroid carcinoma. There are many case reports of unusual presentations of ATC, including bradycardia possibly from compression of the vagus nerve, superior vena cava syndrome, transient thyrotoxicosis, acute Horner syndrome, leukocytosis from secretion of granulocyte colony-stimulating factor, ball-valve type respiratory obstruction, or with primarily gastrointestinal symptoms.^26–33

The aggressive nature of ATC is clearly evident at the time of presentation. Most ATC patients already manifest metastatic disease at this time. Although small, intrathyroidal primary tumors are easiest to treat; only 2% to 9% of patients have such tumors. Most have large tumors with significant extrathyroidal invasion.^15,23 Likewise, from 34% to 62% of patients have local metastases, and from 18% to 46% have distant metastases.^{15–17,22,23} Distant metastases are seen most frequently in lung, around 80% of the time, while bone metastases and brain metastases are seen around 15% of the time.^19,20 All of these reports likely underestimate the dissemination of this disease, accounting for differences in the utilization of diagnostic studies to search for such sites. On those occasions where we have participated in autopsies of ATC patients, metastatic tumors were clearly evident diffusely in sites that were not evident on radiologic staging studies.

Initial assessment of any thyroid mass, particularly one that is rapidly enlarging, entails fine-needle aspiration (FNA) biopsy. A retrospective analysis of 113 ATC patients assessed by FNA biopsies showed that their cytology had demonstrated malignancy in 95%. Analysis of these 107 patients revealed that 90% had been accurately diagnosed as ATC with their biopsy, whereas just under 6% revealed only the coexistent differentiated thyroid cancer and nearly 5% revealed nonspecific malignancy.³⁴ The main reasons for inadequate cytology assessments included tumor regressive changes (necrosis, hemorrhage, and leukocytic infiltration), extensive tumor fibrosis, and sampling errors attributed to the coexistent differentiated thyroid cancer (a problem noted in other reports).³⁵

The gross appearance of ATC describes a tan-white, fleshy, large thyroid tumor that frequently has extrathyroidal infiltration (see Chapter 44, Surgical Pathology of the Thyroid Gland). Portions of the tumor are typically necrotic and hemorrhagic.³⁶ The three predominant histologic patterns seen in ATC include spindle cell in 53%, giant cell in 50%, and squamoid in 19%, with histologic types sometimes overlapping.^37,38 ATC exhibits significant angiotropism (a predilection for growth into the muscular walls of medium-sized veins and arteries),³⁸ cytologic anaplasia showing bizarre tumor cells, and a high mitotic rate³⁶ with a low rate of apoptosis.³⁹ Immunohistochemistry reveals different results in different reported series. Staining for thyroglobulin is seen in anywhere from none to 55% of cases, staining for vimentin ranges from 23% to 94%, cytokeratins from 12% to 80%, both vimentin and cytokeratins are co-expressed in 39% to 75%, and 30% of cases are negative for all of these markers.^{20,37,39–41}

The most recent pathologic staging system, by the American Joint Committee on Cancer⁴² and the International Union Against Cancer,⁴³ denotes all primary ATC tumors as T4, with intrathyroidal primary tumors as T4a and those with any extrathyroidal invasion as T4b. N0 indicates the absence of regional lymph nodes, and N1 indicates the presence of at least one regional metastasis. M0 indicates the absence of distant metastases, whereas M1 indicates their presence. By definition, all ATC cases are stage IV; however, stage IV-A is defined as T4a, any N, and M0. Stage IV-B is defined as T4b, any N, and M0. Stage IV-C reflects the presence of distant metastases (M1), regardless of the extent of the primary tumor or the presence of regional metastases. It is less certain how to classify late transformations of already metastatic differentiated thyroid cancer into ATC; however, any such transformation at a distant site should constitute Stage IV-C.

Careful review of the medical literature prior to 1990 denotes a type of ATC, termed small cell ATC. Since then, this has been found to be a misclassification of tumors that actually were lymphomas, insular carcinomas, medullary carcinomas, or metastatic small cell lung carcinomas.^44,45 For that reason, vintage publications describing series of ATC patients with long-term survivors or significant clinical responsiveness to older chemotherapeutics, such as doxorubicin, should be suspected as having such misclassifications of their patients.^37,46

There are reports of distinctive and unusual variants of ATC. Several cases had acellular fibrotic or infarcted tissue with nests of atypical spindle cells, termed paucicellular variant of ATC, and potentially able to be mistaken for Riedel’s thyroiditis.^47,48 Other cases demonstrate osteosarcomatous differentiation (carcinosarcoma), with prominent osteoid formation and focal calcifications.^49–51 An unusual ATC variant, termed spindle cell squamous carcinoma, seems to be associated with tall cell papillary carcinoma;⁵² however, later reports suggesting similar cases did not report this particular association.^53,54 Since the early 2000s, pathologists have reported a rhabdomyosarcomatous variant of ATC^55–57; however, it is not yet apparent whether such tumors have better or similar behavior to typical ATC. Angiomatoid thyroid tumors have long been reported as variants of ATC; however, it is still unclear whether they represent subtypes of ATC or primary angiosarcomas.^58,59 Likewise, it may be difficult to distinguish a squamoid pattern in ATC from extremely rare primary squamous cell thyroid carcinomas.^60–62

Diagnostic Assessment and Staging

As emphasized earlier, successful management of ATC requires the rapid and effective use of diagnostic and therapeutic modalities. In its most common presentation, as a rapidly growing neck mass, a fine-needle aspiration biopsy should be performed in the office immediately. Although ultrasound guidance may permit the avoidance of biopsies in nondiagnostic necrotic regions, there is no reason to delay the diagnosis and a palpable mass can be easily biopsied without an ultrasound. On the other hand, if an ultrasound is handy, it can be used to guide the biopsy, help evaluate airway patency, and assess for regional metastases. Cytologic results can be obtained within a brief time, permitting a repeat biopsy utilizing flow cytometry and immunocytochemistry if the mass proves suspicious for a lymphoma⁶³ (see the “Thyroid Lymphoma” section in this chapter).

Again, initial staging assessments should be expedited by directly calling the radiologist in order to avoid scheduling delays. Positron emission tomography (PET) with ¹⁸ F-fluorodeoxyglucose (FDG) has proved uniquely valuable in this regard,^64,65 particularly when the image is fused to a coincident computerized tomography scan (FDG-PET/CT).⁶⁶ The cellular basis for the enhanced sensitivity of FDG-PET for ATC is because of its enhanced expression and processing of the glucose transporter, Glut-1, resulting in greater glucose uptake in this tumor phenotype.⁶⁷ Magnetic resonance imaging (MRI) using gadolinium contrast provides the best detail in the neck and superior mediastinum for planning a primary resection in the context of likely infiltrative disease.^68,69 Alternatively, computerized tomography (CT) scans with intravenous and oral contrast can be used if MRI is contraindicated or unavailable. Despite the likely coincident presence of differentiated thyroid cancer, radioiodine scanning and therapy has no role in the acute management of these patients, and concerns regarding stable iodine contamination can be put aside until the ATC has been sufficiently eliminated for a long enough time (usually more than a year) and the underlying differentiated cancer can be addressed. CT scans should be performed from the head through the pelvis in order to completely stage the patient for the presence of distant disease and to document baseline studies.

Overview of Treatment Planning

Effective management of ATC mandates focusing on the appropriate therapeutic target, namely the ATC. Treatment of concomitant differentiated thyroid cancer is put aside until the ATC is effectively treated, as long as differentiated tumor does not involve vital structures or is specifically responsible for significant morbidity. There is no role for radioiodine in the treatment of ATC, and its use has no collateral benefit. Since the early 2000s, many studies have documented that aggressive multimodal therapy (combining surgery, external beam radiation, and chemotherapy) can significantly enhance survival and may even cure some patients, particularly those in stages IV-A and some in IV-B.^70–77 Not all patients, particularly those with extensive distant metastases or with aggressively invasive primary tumors, will benefit from arduous therapies. In addition, the physician should consult the National Cancer Institute clinical trial database (PDQ; http://cancernet.nci.nih.gov/Search/SearchClinicalTrialsAdvanced.aspx) to be informed regarding appropriate clinical trial opportunities.

With all due haste, the multidisciplinary treatment team should be assembled (surgery, oncology, endocrinology, radiation therapy, and palliative care specialists), in person or by teleconference, to consider treatment options for the patient in regard to their disease and clinical status. In some locations, this may take place in the context of a tumor board meeting. A meeting with the patient and their family should follow, providing full disclosure of the risks and benefits of possible treatment options, including innovative therapy, experimental protocols, and palliative care. At the same time, the physician is responsible for fully assessing the patient’s capacity to make decisions. In some cases, this may require consultation with a psychiatrist or a bioethicist. Regardless of the treatment options the patient chooses, he or she should be encouraged to document an advance directive, delineate surrogacy, designate code status, and make end-of-life preferences known.

In the context of the multidisciplinary team, it is important to designate the “captain of the ship,” to provide an easy information conduit to the patient and family and to properly coordinate the specialists’ care. It is easy for these patients to fall between the cracks when dealing with multiple physician practices and their staffs.

Initial Assessment for Surgery

Unlike the typical well-differentiated thyroid cancer patient, the general medical condition of the anaplastic cancer patient is more likely to impact a decision to operate. The demography of advanced age requires a thorough assessment of comorbid conditions that may impact the treatment plan. Surgical contraindications may be relative but may alter planned sequencing of interventions. Hepatic and renal diseases will affect tolerance of adjuvant chemotherapy, whereas a history of previous external beam radiation (XRT) may prevent additional use at that site. Mild dementia frequently blossoms under the stress of cancer therapy. Significantly compromised cognitive function frequently renders delivery of many therapies impractical and potentially unethical. The patient and family need thorough and frank discussions of these issues and must participate in the planning process.

The recurring theme remains, “time is of the essence.” The surgeon must expedite the assessment of readiness for general anesthesia and the stress of surgery. The typical three- to four-week delay to attain imaging studies, preoperative medical consultations and anesthesia consultations, prior to a follow-up office reassessment, is not appropriate. Such a delay to the date of surgery may permit ATC to progress from resectable to unresectable. Endocrine, hepatic, renal, immunologic, neurologic and nutritional issues alone may not contraindicate surgery; however, they must be considered when assessing for an attempt at radical resection. Frequently, a procedure initiated with the goal of total primary resection changes into a debulking procedure. The rapid growth of ATC mandates rapid initiation of adjuvant therapy following surgery. A prolonged convalescence from surgery may compromise the clinical outcome, and should be avoided. Consideration of preoperative (neoadjuvant) XRT may be a reasonable alternative, and may provide sufficient tumor response to permit a primary surgical resection of otherwise unresectable disease.⁷⁸

ATC will typically have potential profound effects on the initial clinical status consequent to dyspnea, dysphagia, vocal cord paralysis, bradycardia, and thoracic outlet obstruction.^3,18 This may require intervention prior to attempted primary resection. Airway preservation and monitoring always have the highest priority because dyspnea—from tracheal involvement, extrinsic compression, or bilateral vocal cord paralysis—needs immediate intervention. The appropriateness of tracheostomy or tumor reduction (with or without airway stenting) needs to be determined and its implementation achieved quickly by competent physicians. Vocal cord paralysis is commonly observed and may be associated with aspiration, particularly if there is concurrent esophageal obstruction below the pharyngoepiglottic segment. Likewise, dysphagia is common, causing dehydration and malnutrition, necessitating enteral access. These issues must be dealt with, concurrent with other diagnostic and therapeutic interventions.

Primary Surgery of Anaplastic Thyroid Cancer

As soon as the diagnosis of anaplastic thyroid cancer is made, the treatment team needs to decide if it is confident that it can surgically manage this disease. Not every surgeon is experienced in the more extensive resections, and possible reconstructive procedures often required for ATC. The adequacy of surgical excision is the strongest predictor of survival,^79,80 and surgeons who are not sufficiently experienced in difficult head and neck cancer surgery may fail to enhance their patient’s survival or provide locoregional control. It is no longer appropriate to exhibit uniform therapeutic nihilism for patients with ATC. If needed, there should be minimal delay in transferring care to an appropriately skilled surgeon. To do otherwise may allow sufficient tumor progression so as to negate all therapeutic intentions. Worse, a surgical salvage operation, after an initial resection failure by a less experienced surgeon, is associated with treatment delay, contamination of tissue planes, possible tumor implantation, and accelerated tumor growth.⁸¹