There are only a limited number of publications about prognostic factors for disease-specific survival in patients with HCTC from different single institutions [1, 2, 6, 7, 9, 11, 14, 15, 17, 19, 23]. In these publications, 20 different prognostic factors were reported. Prognostic factors were related to: patient’s characteristics [6, 19], tumor [6, 9, 11, 17, 19, 23], extent of disease [1, 6, 9, 17, 19, 23] and treatment [1, 6, 19, 23]. The majority of series included only a very limited number of patients, therefore the authors of these studies could perform only univariate statistical analysis. To our knowledge, so far only six publications [1, 2, 6, 11, 19, 23] have reported the results of multivariate regression analysis of prognostic factors for disease-specific survival in patients with HCTC. The following prognostic factors were reported: age of patients [2, 6], gender , marital status , tumor size [2, 6, 17], extrathyroidal invasion [2, 6, 19], tumor stage , regional metastases , distant metastases [6, 19], capsular invasion , multifocal disease , solid pattern of growth , and extent of surgery . In our patients, the independent prognostic factors for disease-specific survival were: age of patients, distant metastases, and residual tumor after surgery. Two of the previous six papers were population-based [2, 6], while two were clinicopathologic [11, 17]. To continue, two papers reported data from major cancer comprehensive centers covering large populations [19, 23]. However, both of them included a smaller number of patients than our study, therefore our paper provides more robust data on a larger number of prognostic factors.
The mean age of our patients at diagnosis was 58 years, which is in accordance with the majority of reports in the literature [1, 6, 10, 14, 15, 23]. Our results confirm the findings in the literature [2, 6, 17] and show that patients’ age is an independent prognostic factor for both disease-specific survival and disease-free interval. In patients younger than 45 years of age, shorter disease-free survival was as many as 17.04 times more likely than in older patients.
Distant metastases were an independent prognostic factor for disease-specific survival in our patients. In patients with distant metastases, the risk of shorter survival was 3.3 times higher than in patients without distant dissemination. This relative risk was 5.3 in the study of Lopez-Penabad et al. .
Residual tumor after surgery was another independent prognostic factor for disease-specific survival in our population of patients. Our multivariate analysis also included the following confounding factors related to locoregional tumor extension and treatment: tumor size, tumor stage, regional metastases, capsular invasion, extrathyroidal invasion, and extent of surgery.
The recurrence rate in our patients was 27%, which is comparable to reports in the literature [1, 17, 19, 23]. Khafif et al. reported that locoregional recurrence occurred in only 4 of 42 patients (10.5%) . The probable cause for their low recurrence rate is the obvious difference in patient selection between our and their study groups. Their patients received primarily private consultations with only two surgeons. Therefore, HCTC was usually diagnosed at an earlier stage . On the other hand, our patients had a lower recurrence rate compared to patients from another two tertiary cancer comprehensive centers: Royal Marsden Hospital and Sloan-Kettering Cancer Center, with the recurrence rate of 34% and 43%, respectively [23, 17]. Possibly, the lower recurrence rate in our patients was related to more frequent use of adjuvant multimodal therapy.
The 10-year disease-free survival was 68% in our patients, while it was reported to be 40.5%, 43%, and 61% by Kushcayeva et al. , Mills et al. , and Stojadinovic et al. , respectively. To our knowledge, only six reports [1, 14, 15, 17, 22, 23] were published about disease-free survival and 15 factors were found to correlate with recurrence of HCTC. Predictive factors were related to: patient’s characteristics , tumor [14, 15, 17, 22], extent of disease [1, 15, 17, 23] and treatment [1, 23]. However, the majority of previous papers had a small number of patients, and only two publications [15, 23] reported the results of multivariate logistic regression analysis of the disease-free interval in patients with HCTC. Our multivariate analysis showed that independent prognostic factors for disease-free survival were: gender, age of patients, regional metastases, and residual tumor after surgery. Regional metastases have already been a known independent prognostic factor from the study of Stojadinovic et al. . Extrathyroidal invasion was another independent prognostic factor in their study . However, in our study group, extrathyroidal invasion was a prognostic factor by univariate analysis only and was not an independent factor, while residual tumor after surgery was an independent prognostic factor. Furthermore, an old clinical observation that younger patients and females have a favorable prognosis was statistically proven by our results.
The 10-year disease-specific survival in our patients was 88%, while it was 49%, 64%, and 73% in studies reported by Kushcayeva et al. , Mills et al. , and Stojadinovic et al. , respectively. It should be stressed that survival rates represent the result of both selection bias, i.e. more advanced disease in larger tertiary centers, and the effectiveness of different treatment modalities used in patients. Fortunately, SEER  and our data show that nowadays the majority of patients with HCTC have a favorable prognosis. Total thyroidectomy and RAI ablation of thyroid remnant tissue enable early detection of tumor recurrence. Furthermore, it is possible that such treatment affects disease-free and disease-specific survival. In addition, our results also show that long-term survival can be obtained also in patients with locoregionally advanced and metastatic disease, if they are treated multidisciplinary. Chemotherapy before surgical procedure may be effective in order to decrease the tumor size in HCTC . Our group reported that RAI therapy may be effective in patients with metastatic HCTC  and that recombinant human thyrotropin-aided RAI therapy may be effective in patients with metastatic HCTC . Since residual tumor after thyroid surgery is an independent prognostic factor for disease-specific as well as disease-free survival, an effective locoregional therapy is mandatory. Therefore, EBRT should be used in cases of residual tumor.
The limitation of our study is that the results of treatment and survival are reported only from a single institution. Another limitation is that not all our patients were treated uniformly because during a 40-year time frame, they were treated by many surgeons and oncologists. Our study is observational and not randomized, thus it is not possible to draw conclusions about the impact of treatment on patients' survival. However, the advantage of our study is that it included all patients with HCTC from Slovenia treated in the period from 1972 to 2011. During this period, all patients with thyroid carcinoma were referred to the only tertiary referral center for postoperative treatment and follow-up of patients with thyroid carcinoma in our country. Thus, our data are more comparable to the study reported by Goffredo et al. , a population-level analysis of 3,311 patients from the Surveillance, Epidemiology, and End Results (SEER) database from 1988 to 2009 in the USA, than to reports from cancer comprehensive centers. Patients with more advanced disease are more often treated in these centers than in community hospitals [9, 11, 15, 17, 19, 22–24]. Goffredo et al.  reported that the female-to-male ratio was 3.22:1, mean age of patients was 58 years, mean tumor size was 3.6 cm, and extrathyroid extension was present in 14% of cases. Initially, only 5% of their patients had regional and 5% distant metastases. However, our patients had more advanced disease: mean tumor size was 5 cm, extrathyroid extension was present in 22% of cases, 7% of patients had regional and 11% distant metastases. Probably, our patients had more advanced disease because Slovenia is an endemic goiter region . However, our patients had similar 10-year disease-specific survival compared to the SEER database patients reported by Goffredo et al. . This can be explained by more effective treatment of our patients. Total thyroidectomy enables more effective adjuvant treatment with different modalities (radioactive iodine, EBRT and/or chemotherapy) resulting in longer survival, which is particularly important in patients with more advanced HCTC. Total thyroidectomy, RAI ablation of the thyroid remnant, and EBRT were done in Goffredo’s  and our patients in 75%, 45% and 4%, and 71%, 87% and 27%, respectively.