Analysis on The CSS Difference and Its Inuence Factors Between ATC and PSCCT

Backgrond:The clinical manifestations and treatments of ATC and PSCCT are similar .Their prognosis is worse than other types of thyroid cancer. This study aims to analyze on the CSS Difference and Its prognosis Factors Between ATC and PSCCT. Methods:We reviewed the data of patients who were diagnosed as ATC and PSCCT between 2004 and 2015 in the Surveillance, Epidemiology, and End Results (SEER) database. Statistical analysis included Kaplan-Mier survival analysis, Cox regression models and the log-rank test. The predictive accuracy ability of the model were determined by concordance index (C-index) Result:1164 patients with ATC and 124 patients with PSCCT were included. The clinical treatment rate and distant metastasis rate of ATC is higher than that of PSCCT(p<0.05). PSCCT CSS is better than ATC CSS (cid:0) p=0.0074 (cid:0) . Multivariable analyses showed that Age,Sex, Radiation,Chemotherapy,Sequencenumber,Marital,SurgPrimSite or M stage were independent prognostic factors,the C-index of this model is 0.803 (95%CI: 0.787-0.819). Age, Radiation, Sequencenumber, SurgPrimSite are independent prognostic factors of PSCCT CSS, C-index of the model is 0.829, (95%CI: 0.786-0.872) Conclusions:Although the overall CSS prognosis of PSCCT patients is better than that of ATC patients, the prognostic factors of ATC and PSCCT are not exactly the same. It is necessary to formulate clinical treatment and management plans based on the type of disease. as CSS) of ATC and PSCCT. We evaluated the CSS of the two types of thyroid cancer with Kaplan-meier method and evaluated the difference of the CSS with Log-rank method. Single-factor and multi-factor COX regression were used to evaluate the impact of demographic and clinical indicators on the CSS of ATC and PSCCT. Harrel’s concordance was used to evaluate research model. We adopted the level of two-sided 0.05 for all p-values.


Introduction
Anaplastic thyroid carcinoma (hereinafter referred to as ATC) is the most aggressive thyroid tumor among primary thyroid cancers, which accounts for about 1-2% of the total and occurs mostly in the elderly. There is no clear differentiation in anaplastic, while giant, spindle and squamous cells show obvious pleomorphism 1,2 . Primary squamous cell carcinoma of the thyroid(hereinafter referred to as PSCCT) is an extremely rare and highly aggressive malignant tumor, accounting for less than 1% of all primary thyroid cancers and resulting in unfavourable prognosis. In terms of cytology, it is only composed of large and pleomorphic epithelial cells with keratinization, and necrotic components are common-seen 3,4 . The most common symptom of PSCCT is an enlarged anterior neck mass, followed by di culty breathing or swallowing and changing voice. ATC and PSCCT both show abundant squamous cell-like cells. Therefore, they are easily confused in clinical diagnosis. Related studies showed that they are identical in clinical manifestations for the similar treatment is adopted.
In fact, the research on the difference and correlation between ATC and PSCCT has drawn an increasing attention. However, it is di cult to carry out prospective studies based on large samples due to the rarity. With poor prognosis, these two carcinomas are highly aggressive for human body. Nowadays, though combined treatments including surgery, radiotherapy and chemotherapy are often adopted in clinical treatment 6 , different studies have gured out that adjuvant therapies exert unclear impact on them 7,8 . However, the difference of CSS between ATC and PSCCT has not been extensively studied, and it is highly necessary to study the prognostic factors of CSS between ATC and PSCCT.
In order to better understand the demographic and clinical treatment-related indicators of ATC and PSCCT so as to have a difference analysis and explore the prognosis of CSS and its in uencing factors, we conducted a retrospective study based on the Surveillance, Epidemiology, and End Results Program (SEER) 18-registry database 9 .

Data Source
The study conducted an in-depth analysis on the registered data of The Surveillance, Epidemiology, and End Results Program database and extract data from 18 Regs Custorm Data (with additional treatment feilds).Nov 2018 Sub (1975-2016 varying) with SEER*Stat 8.3.6. The patient information in this data set has been removed, so this procedure does not involve patient's consent.

Study Population
The carcinoma samples adopted in the study are ATC or PSCCT diagnosed betwen 2004-2015. We selected samples based on the organizational behavior code in the International Classi cation of Diseases for Oncology third edition (ICD-O-3) issued by the World Health Organization (WHO) in 2008, in which the histology code of ATC is 8012, 8020, 8021, 8030 -8032, the histological code of PSCCT is 8070-8076, and the primary thyroid location code is C73.9 for all samples. The study included demographic and clinical data related variables, such as the year of diagnosis, patience's age, race, gender, and marital status at diagnosis, whether or not multiple primary carcinoma, surgical status, N stage, M stage, radiotherapy, chemotherapy, survival time, and SEER cause-speci c death classi cation and excluded data where the race survival time was unknown, the surgical status and N stage are Blanks.

Statistical analysis
We adopted chi-square test to analyze count data rate or percentage difference. The destination of this study is the cancer-speci c survival (hereinafter referred to as CSS) of ATC and PSCCT. We evaluated the CSS of the two types of thyroid cancer with Kaplan-meier method and evaluated the difference of the CSS with Log-rank method. Single-factor and multi-factor COX regression were used to evaluate the impact of demographic and clinical indicators on the CSS of ATC and PSCCT. Harrel's concordance index (C-index) was used to evaluate research model. We adopted the level of two-sided 0.05 for all p-values.
The drawing and statistical calculations in this study were all carried out by R language 3.6.0 (http://www.r-project.org/).

Results
Analysis on the difference between demographic and clinical indicators By selecting, this study included 1288 patients with primary thyroid cancer in SEER database from 2004 to 2015, in which there are 1164 patients with ATC and 124 patients with PSCCT. The difference analysis of demographic and clinical data of ATC and PSCCT is shown in Table 1. The study gured that the prevalence of ATC and PSCCT is mostly concentrated in the elderly, in which the prevalence of females is higher than that of males. However, there is no difference in the distribution of age, gender, race, marital status, and proximal metastasis between the two diseases. Compared with PSCCT patients, ATC patients have a higher proportion of chemotherapy, radiotherapy, surgical conditions, distant metastasis, and multiple primary conditions, thus the difference was statistically signi cant (p<0.05).  10 . It also con rmed that ATC has a higher proportion of distant metastases, which is in line with the study conducted by scholars like Glaser.
Consistent with previous research results, the proportion of women with ATC and PSCCT is higher than that of men, which may be related to the biological characteristics of the cancer itself 12,13 . Because PSCCT is a rare histological type of thyroid cancer and the clinical treatment experience is limited, it's treatment is controversial. At present, the treatment methods of PSCCT and ATC are similar, and most of them are chemotherapy, radiotherapy, and surgery. However, the treatment rate of ATC is higher than that of PSCCT. The reason is that compared with PSCCT, ATC has clear histology and positioning 10 .
Age is an in uencing factor of the CSS of ATC patients, for example the CSS of patients older than 65 years old is less than that of patients younger than 65 years old, which has been con rmed by the study conducted by researchers like Gui. Studies conducted by Smallridge and other scholars have proven that for early patients diagnosed with ATC, surgical treatment, chemotherapy, radiotherapy and systemic treatment can achieve the optimal survival goals for patients, and active palliative care and clinical care for patients with terminal ATC are also important components for achieving the optimal survival goals 16 .
Relevant studies 17, 18 have con rmed that gender, surgical treatment, radiotherapy and chemotherapy are all prognostic factors of the CSS of ATC patients, which are consistent with the results of this study. This study showed that marital status is an in uencing factor for the CSS of ATC patients. Many studies have suggested that various family care and psychological care in different marital status cause different survival conditions of patients 19,20 .
Our study con rms that PSCCT is more likely to occur in elderly female patients since age is an independent factor, but gender and race are not. Radiotherapy is a protective factor for the CSS of PSCCT patients, which is echoed by related studies 21 . Shimaoka's 21 case study suggested that chemotherapy plays an active role in the treatment of PSCCT, but in this study chemotherapy is not an in uencing factor for the CSS of PSCCT patients. Surgical treatment is also an independent factor for the CSS of PSCCT patients. Studies gured that the best treatment for PSCCT is early diagnosis and then aggressive surgery to prevent bleeding or suffocation caused by obstruction 23 . Different from the CSS of ATC patients, marital status is not an in uencing factor for the CSS of PSCCT patients. Studies conducted by scholars like AU 7 found that surgery, radiotherapy, and combined surgery and radiotherapy are not in uencing factors for the CSS of PSCCT patients. This is different from the results of this study, which may have something to do with the surgical and radiation method. This study shows that surgical treatment is a protective factor for the CSS of PSCCT patients. In this study, the distant metastasis variables in the model were statistically tested to close the limits. The HR value of multiple primary cases is larger, which may be caused by limited amount of samples in addition to the actual effect of the variables.
By the comparison between the survival rates of patients with single primary and multiple primary cancers, Amer 24 found that the overall survival rate of patients with multiple primary cancers is higher than that of single primary cancers. This may be related to the characteristics of thyroid cancer for the survival time of many types of cancer patients is not long enough to develop to the second origin. Studies conducted by researchers like Lin 17 showed that although the OS of patients with multiple primary thyroid cancer is much lower than that of single primary thyroid cancer, there is no difference between their CSS. This study gured that multiple primary cancers are a protective factor for the CSS of patients with ATC and PSCCT, which does not mean that the mortality rate of single primary thyroid cancer is lower. This may be because the mortality rate of other cancers in multiple primary patients is higher than that caused by thyroid cancer.
This is the rst study on CSS of ATC and PSCCT patients based on a large population cohort. Both ATC and PSCCT are aggressive thyroid malignancies, and the CSS prognosis of PSCCT patients is favourable than that of ATC patients. We have established prognostic models of CSS for patients with ATC and PSCCT respectively. The prognostic factors of the two are not exactly the same, which means although the histological manifestations of ATC and PSCCT are similar to the current treatment plan, the clinical treatment and management plans need to be targeted for different type of diseases.

Declarations
This study was supported by funds from the National Natural Science Foundation of China (61803112, 31660320), the Science and Technology Ethics approval and consent to participate Not applicable.

Consent for publication
Not applicable.

Availability of data and materials
Study data was publicly available in the SEER database.

Competing interests
The authors declare that they have no competing interest. In order to study the CSS prognosis of ATC and PSCCT, we visually presented it and analyzed the difference. After drawing the Kaplan-Meier curve of the study samples, the median survival time of CSS of ATC patients was 4 months, and PSCCT patients was about 10 months. After Log-rank test, it is shown that CSS of PSCCT patients was higher than that of ATC patients. Therefore, the difference is statistically signi cant (p<0.01) Figure 1).