Our results showed the majority of ATC were older than 60 years (73.6%) with a male to female ratio of 1:1.9, very similar to previous findings 6–9. However, most of our ATC patients (72.7%) presented with tumors ≥ 5 cm in diameters slightly larger than in previous studies (53.0–68.9%)6,10. According to TNM staging, our patients were approximately equally distributed across each staging (35.2% IVa, 37.1% IVb, and 27.6% IVc). In the case of stage IVb and IVc cases with extra-thyroid invasion, thyroid gland surrounding tissue was frequently involved making complete removal difficult in patients with extensive involvement of vital structures. We found that 11.6% of our patients presented with tumors involving vital structures. Distant metastasis was also a poor survival outcome prognostic factor found in 22.6% in our series. Previous research reported ATC survival outcomes ranging from 2–10 months and > 2-year survival rate of 0–10%11, that similar to our findings of median survival time and 1-year survival rate of 77–82 days and 3.5-4.0%, respectively. Survival outcomes in prior retrospective studies vary depending on sample size, baseline demographic data, and selection bias. Our survival outcomes were likely worse than previous studies due to numerous cases with a huge tumor, involving vital structures, and distant metastasis. Treatment modality has important effects on survival outcomes. Some studies advocate multimodality treatment has benefits12–14; however, few studies report significant survival benefit from multimodality treatment15,16. Our study found that intervention treatments provided better survival outcomes than palliative treatment (p < 0.05) in overall staging. However, among the intervention treatment groups, the surgery and postoperative chemoradiation combination provided the best 1-year survival rate of 20.0%. These findings compare well with a previous study14 that showed complete ATC resection combined with postoperative adjuvant chemotherapy and irradiation resulted in longer-term survival, even with persistent minimal disease. Although intervention treatment overall seemed to provide superior survival outcome benefit, we were also investigated possible differential effects across different staging levels. We found intervention provided significantly better outcomes over palliative care in stage IVa (p < 0.05). Intervention treatment was also better than palliative care in stage IVb and IVc (p > 0.05) but not at a statistically significant level possibly due to more aggressive tumors in these advanced stages.
Age, gender, tumor size, the extent of disease at presentation, acute symptoms, distant metastasis, leukocytosis, and multimodality therapy) are previously reported prognostic variables for survival outcome 13,14,17−21. In our study, the ATC patients’ prognosis mainly depended on age, leukocytosis, and treatment. Glaser et al22 reported that age ≥ 65 years was an unfavorable prognostic factor. This finding was similar to our study that showed older age as the significant higher mortality factor (HR of 1.55). Other authors had also reported that older age was a poor prognosis factor but old age was variously defined. Old age range was reported as ≥ 60–75 years in the previous studies20,23,24. Furthermore, leukocytosis was observed that also predicted poor survival outcomes. Jiang et al6 and Sugitani et al25 found the HR of 1.12 and 1.48, respectively. In our series, a white blood cell count ≥ 10,000/ml3 was analyzed with Cox regression that revealed a hazard ratio of 2.76 (p < 0.001). This finding was comparable with previous reports investigating effects of leukemoid paraneoplastic reaction by ATC tumor secreted cytokines including granulocyte-colony stimulating factor, granulocyte macrophage-CSF, and interleukin-626,27. The last significant prognostic factor in our study, treatment modality, revealed that palliative treatment predicted the poorest overall survival outcome with a HR of 1.85 (p < 0.05). However, selection bias makes this finding unsurprising given that palliative care patients usually had advanced disease with high mortality.
Sugitani et al25 classified the modality benefits in each ATC staging and found post-operative chemoradiation was a significantly favorable prognostic factor in stage IVb (HR of 0.45; p = 0.083) whereas at stage IVa its benefits did not reach a statistically significant level (HR of 0.21; p = 0.19). Although there is still controversy about proper ATC treatment protocol, several previous studies suggest multimodal treatment allows a longer ATC survival rate. Kobayashi et al19 suggested active multimodality for the early stage. The multimodality protocol of surgery and chemoradiation has been advocated as offering the longest survival rate28–30. In the present study, the combined modality of postoperative chemoradiation and radiotherapy offered a longer median survival time of 187 days and 177 days, respectively, over surgery alone, which provided a survival rate of 64 days, again supporting a multi-modality advantage in the survival outcomes. A negative prognostic effect has been reported for hypothyroidism. Our study found hypothyroidism was a negative predictor with using univariate analysis but the multivariate regression model analysis showed that this difference was not statistically significant. Jiang et al6 found similar findings where serum T4 level was not found statistically significant in Cox regression analysis. However, they observed that patients with low T4 level had significantly lower survival rates than those with normal T4 levels. Several authors proposed that hypothyroidism may be occurring via tumor destruction of the normal thyroid tissue31,32 and inhibition of changing of T4 to T4 binding globulin by unsaturated fatty acid from hypoxic or injured tissue in severe illness patients33. Therefore, low T4 levels may represent a late stage of ATC with severe disease that indicates poor survival outcomes.
ATC is an extremely aggressive rapidly progressing tumor that makes it difficult to use a randomized prospective protocol for evaluation of treatment and survival outcomes; therefore, a retrospective chart review was selected on feasibility basis for this study. Although our study includes the limitations of retrospective studies, it did show multi-modality treatment was superior to palliative modality, especially the combination of surgery and chemoradiation. Furthermore, we found that not only palliative treatment but also age and leukocytosis were unfavorable prognostic factors for predicting mortality outcomes. In future, more laboratory information and detailed clinical data would allow for better investigation of prognostic factors.