Anaplastic thyroid carcinoma is a highly lethal cancer with a spontaneous median survival of 4 to 12 months from the time of diagnosis . It represents less than 2% of all thyroid cancers but still accounts for up to 40% of thyroid cancer mortality [9, 10].
The characteristics of the patients in our study were similar to those described in the literature with an age between the sixth and seventh decades. There was a slight female to male predominance of 1.5 to 1. A large percentage of patients already had distant metastases at presentation. As in our study, the most common site of metastases reported in a 50-year review by McIver et al.  was the lung (42%), followed by bone (32%) and brain (9%). Although the identification of histological subtypes is an important part of the diagnosis of ATC, histological characteristics do not provide prognostic or predictive information regarding the patients' disease course or response to therapy. Because of a threat of local or metastatic spread, therapeutic options in ATC must combine local (cervical) and systemic treatments. Single modality therapy has been shown to have limited effect. Surgery or radiation, or chemotherapy alone rarely have significant impact on local-regional disease recurrence and overall survival. In the study by Veness et al. , 18 patients with ATC were treated at a single center, with 9 having single modality treatment (6 had radiation therapy, 3 had surgery, and 1 had chemotherapy). Four of the six patients receiving radiation therapy and two of the three patients undergoing surgery died of uncontrolled local disease. Because of the disappointing results from single modality therapies, multimodality therapy has progressively become the treatment of choice in ATC, with surgery being the cornerstone of patient management. Nearly all long-term survivors with ATC have had surgery as part of their treatment. In a retrospective review, Haigh et al.  describe five longterm survivors treated with surgery followed by RT/chemotherapy. More specifically, those patients who underwent a potentially curative resection (8/26) had a median survival of 43 months compared to 3 months for those with a palliative resection. In the study by Swaak-Kragten et al , patients who had R0/R1 resection and underwent protocol chemoradiation had a local complete response rate of 89%, compared to only 3% CR for patients receiving neither R0/R1 resection nor protocol radiation. Complete locoregional response resulted in longer median survival (7 months versus 3 months) and improved 1-year overall survival (32% versus 9%). Levendag et al.  has also described the achievement of a CR as an important goal for treatment. The above data confirm both the central role of surgery as part of multimodality therapy for ATC and the potential for long-term survival in a subset of patients. Unfortunately, as seen in our study and as reported by Haigh et al. , only a minority of patients can undergo microscopic tumor resection at the time of presentation.
The history of RT in the treatment of ATC has evolved from high dose conventional RT to hyperfractionated RT administered in combination with continuous chemotherapy. Since ATC is a rapidly dividing tumor, hyper-fractionated RT minimizes the opportunity for tumor cells to recover between treatments. Tennvall et al.  have reported improvement in local control with acceleration of hyperfractionated RT and concurrent chemotherapy. All 33 patients who achieved local control had total resection, but 30 had microscopic residual disease. This supports the hypothesis that RT can eradicate microscopic disease and improve survival when given in effective doses and fractionation schemes. Results published by Swaak-Kragten et al.  show that the total radiation dose appears to be an important prognosis factor; median overall survival was 5.4 months for patients treated with a total dose of > 40 Gy versus only 1.7 months for those receiving < 40 Gy. In the same study, the authors have added prophylactic lung irradiation (PLI) in combination with low-dose doxorubicin because of the very high percentage of patients developing lung metastases in the course of their disease. This treatment had not been described in the literature for ATC before. Interestingly, of the 11 M0 patients treated only two subsequently developed lung metastases; however, a significant benefit could not be demonstrated due to the small number of patients treated and the lack of a control arm. Some chemotherapy drugs have been tested to try to increase overall survival in metastatic ATC patients. Results of a randomized trial by Shimaoka et al.  comparing doxorubicin alone to doxorubicin and cisplatin in 39 patients with ATC have shown a response rate of 5% for the single agent compared to 18% for the combination therapy. The authors have concluded that combination therapy is superior in terms of efficacy but this efficacy does not balance the risk of increased toxicity. Ain et al.  have employed paclitaxel in a phase II trial and have demonstrated that treatment responders had a median survival of 32 weeks compared to only 7 weeks in non-responders. Multimodality treatments have demonstrated efficacy in improving local tumor control in some ATC patients and may increase survival in patients with good performance status and without renal or cardiac insufficiency or metastatic disease. Unfortunately, most of the studies published were retrospective or were based on very small numbers of patients [18, 19]. In recent years, two important prospective trials have been published. First, Tennvall et al.  have treated 55 consecutive patients with three different treatment schedules, all containing surgery, hyperfractionated radiotherapy and chemotherapy.
Despite this aggressive approach, overall survival was poor with only 9% of patients surviving at 2 years. A second study by the IGR team, published by De Crevoisier et al. , has enrolled 30 patients (including 6 with lung metastases) treated with surgery, chemotherapy (doxorubicin and cisplatin) and hyperfractionated RT. Results described a survival of 46% at 1 year and 27% at 3 years. These results are similar to ours (40% and 25%, respectively) but higher than in most other studies [5, 18, 19]. When considering our patients and those from the IGR study we obtain a total of 74 French patients treated with the same protocol. The results of the 2 studies are equivalent, justifying the use of a combined treatment in different cases of AT, principally for non metastatic patients after thyroid surgery (either complete or incomplete), whatever the lymph node status (N0/N1). Secondly, for good PS metastatic patients with limited metastatic lesions we would recommend a multimodality treatment (chemoradiotherapy) and then total thyroidectomy whenever feasible. It clearly appears that the primary step of clinical management in ATC is to identify potentially curable patients with ≤3 factors of poor prognosis and to offer them multimodality treatment to improve their chances of survival. Regarding patients with more than 3 poor prognosis factors, we should avoid aggressive therapy and offer them best supportive care.
Regarding new antitumor therapies, ATC have poorly benefited from recent progresses. Some have been tested in ATC but results have been often disappointing. Targeted agents directed against VEGF-R like axitinib (AG-013736) or against BRAF and VEGF-R like sorafenib (Nexavar®) have been tested in phase II trials including patients with anaplastic thyroid carcinoma [20, 21]. One response was obtained with axitinib and one stabilization with sorafenib.
Vascular disrupting agents as Zybrestat® (fosbretabulin, combretastatin A4 phosphate (CA4P)) have been evaluated in ATC because of their antiangiogenic and cytotoxic activity. The promising results obtained in vitro foreshadowed the clinical efficacy of the drug . at the 35th European Society of Medical Oncology (ESMO) meeting, the preliminary results of a phase II-III study of Zybrestat® in combination with chemotherapy in ATC were presented. In this study, 80 patients (out of the 180 planned) were randomized to receive Zybrestat® (IV once a week) in combination with chemotherapy (paclitaxel-carboplatin every 3 weeks) versus chemotherapy alone. For the first time in ATC, the median overall survival time was increased by the association (5.1 months versus 4.1 months for patients receiving chemotherapy alone).