Prognostic value of preoperative chemotherapy for thymic epithelial tumors: A propensity-matched analysis based on the SEER database

Background The aim of this study was to assess the impact of preoperative chemotherapy on long-term survival (≥1 month) in patients with thymic epithelial tumors (TETs) and conditions suitable for chemotherapy using data from surveillance, epidemiology, and end-result databases. Methods This retrospective study controlled for confounding factors by propensity score matching (PSM), analyzed overall survival (OS) and cancer-specific survival (CSS) by Kaplan-Meier methods, and analyzed factors affecting the prognosis of patients undergoing surgery for thymic epithelial tumors by univariate and multifactorial Cox regression. Results A total of 2,451 patients who underwent surgery for TETs were identified from the Surveillance, Epidemiology, and End Results database. Preoperative chemotherapy significantly improved OS and CSS in patients with stage III/IV TETs compared to patients without preoperative chemotherapy. Subgroup analysis showed that patients younger than 60 years of age with TETs, patients with thymic carcinoma, and patients with TETs with multiple cancers were more likely to benefit from preoperative chemotherapy. Conclusion This study found that preoperative chemotherapy is a viable option for advanced thymoma with favorable overall and cancer-specific survival rates, but patient history and physical condition should be fully considered in conjunction with diagnostic imaging findings to assess patient tolerance to chemotherapy.


Introduction
Thymic epithelial tumors (TETs) represent a group of heterogeneous, rare neoplasms arising from thymic epithelial cells and are the most common tumors of the anterior mediastinum (1). With disease progression, neoplastic cells invade mediastinal and thoracic organs, such as the lungs, heart, great vessels, surrounding nerves and lymph nodes, and may damage those organs (2). Thymic epithelial tumors include thymomas Abbreviations TETs, thymic epithelial tumors; POCT, preoperative chemotherapy; OS, overall survival; CSS, cancer-specific survival;SEER: Surveillance, Epidemiology, and End Results; Ts, thymoma, TCs, thymic carcinomas; NENTs, thymic neuroendocrine neoplasms; K-M, Kaplan-Meier.
The rarity of TETs, with an overall incidence of 0.13-0.32 per 100,000 people per year, has somewhat limited prospective studies, and optimal treatment options remain an unresolved issue (6). Studies on the effects of preoperative chemotherapy in patients with TETs are still inadequate, and the prognostic impact of chemotherapy on patients with TETs is still controversial and requires further study (7)(8)(9)(10)(11).
In this study, we aimed to assess the prognostic value of preoperative chemotherapy for TETs, its safety and its optimal conditions of application.

Patient selection
The Surveillance, Epidemiology, and End Results (SEER) database is one of the largest publicly available databases, with approximately 28% of the U.S. population covered (12). In this study, all cases were obtained from the SEER program (www. seer.cancer.gov), and patients from the SEER  We identified 3,555 cases according to the following admission criteria: (1) year of diagnosis from 2007 to 2019; (2) ICD-O-3 site code C37.9 (thymus); (3) pathologically confirmed TETs, not diagnosed by autopsy or death certificate; (4) patients with complete data on age at diagnosis, sex, stage, treatment, histology, vital status, or months of survival; (5) considering that surgery may lead to immediate death in the short term (16) and the lack of information on postoperative complications in the SEER database, patients surviving <1 month were excluded from this study, which focused on the impact of the treatment approach on long-term survival (≥1 month) in patients with thymic epithelial tumors.
For further study, patients who met the following criteria were excluded: (1) no/unknown cancer-directed surgery of primary site performed; (2) unknown sequence of surgery and radiotherapy; (3) the systemic treatment received was not chemotherapy; (4) age at diagnosis less than 18; (4) ethnicity information unknown Figure 1 demonstrates the flowchart of case inclusion and exclusion in detail.
This study used previously collected anonymized and deidentified data from the SEER database. Therefore, this study was exempted from ethical review by the Institutional Review Board of the Fifth Affiliated Hospital of Zhengzhou University.

Study Variables
The variables involved in our study included basic demographic information (age at diagnosis, sex, marital status, and race), neoplasm-related information: tumor size, WHO classification (A, AB, B1, B2, B3, TCs and NETs), Masaoka-Koga Stage (I/IIA, IIB, III/IV), Number of tumors (One primary tumor only, With other malignant tumors), and therapeutic information: Lymph node biopsy (Negative, Not performed, Positive); Radiotherapy Information (no_Radiotherapy, Radiotherapy after surgery, Radiotherapy prior to surgery); Chemotherapy (no_systematic treatment, Preoperative systemic treatment, Postoperative systemic treatment), survival information: survival months (from diagnosis to death or last follow-up), vital status (Live, Dead). Overall survival (OS) and cancer-specific survival (CSS) were the primary study endpoints. OS was defined as the time from diagnosis of TETs to death or loss to follow-up for any reason; CSS was defined as the time from the date of diagnosis to direct or indirect death from thymic epithelial tumors.
For two groups of continuity indicators (age and tumor size) X-tile software was used to select the best cut-off point in the survival data and to group the age and tumor size. The optimal groupings for age at diagnosis in this study were the <60 years group and the ≥60 years group; the optimal groupings for tumor size were the <80 mm group, the ≥80 mm group and the unknown size group. For marital status, patients were divided into a married group, a single (never married) group, and an other group. For race, patients were grouped into white group, black group, and Asian/Other ethnic group (including Pacific Islander, Alaska Native, etc.). The sequence of surgical vs. systemic treatment is recorded in the SEER database, in which systemic treatment mainly refers to chemotherapy, but also includes hormonal treatment, BRM treatment and transplant/endocrine cases. In this study, we included only cases where the type of systemic treatment was chemotherapy. The staging in the SEER database was divided into local, regional, and distal disease, and we reclassified the included patients with TETs into three groups according to the corresponding Masaoka-Koga staging as follows (the exact correspondence is indicated in Table 1): stage-I/IIA (Localized only), stage-IIB (Regional), and stage-III/IV (Distant site(s)/ node(s) involved) (11,17).

Statistical analysis
This study was analyzed using R statistical software (www. r-project.org). Among patients who underwent surgery for TETs, Pearson χ 2 tests or Fisher's exact tests were performed for patients who received different treatment modalities (no preoperative chemotherapy and preoperative chemotherapy) and treatment-related factors. Univariate and multifactorial Cox regression models were performed using "tableone", "dplyr", and "skimr" in R software to estimate hazard ratios (HR) and 95% confidence intervals (CI) to analyze independent prognostic factors associated with overall survival (OS) and cancer-specific

Definition of Masaoka-Koga stage
The staging in the SEER database (17,18) I: Completely encapsulated tumors in the gross and microscopic "Localized only": "Invasive carcinoma confined to the primary gland" or "localized, not other specified" IIA: Percutaneous invasion under microscope IIB: Gross invasion of the thymus or adjacent adipose tissue, or severe adhesions without breach of the mediastinal pleura or pericardium Frontiers in Surgery survival (CSS) for patients undergoing TETs. Kaplan-Meier curves were plotted using the "Survival" package and "ggsurvplot" in R software to estimate OS and CSS for each group of patients, and P values were determined using the log-rank method. A 1 : 1 optimal nearest neighbor propensity score matching (PSM) was performed using the "MatchIt" package in R software to balance the baseline characteristics of patients in the study and control groups with a caliper value of 0.1. P < 0.05 was considered a statistically significant difference.

Distribution Characteristic of Factors Related to Treatment Patterns
Our study enrolled 2,451 eligible patients who underwent surgery for TETs between 2007 and 2019. Among them, 1,686 were white, 346 were black, and 419 were patients of other ethnicities, including Pacific Islanders and Asians; The study TABLE 2 Clinicopathological characteristics of patients with thymic epithelial tumor surgery before and after propensity score matching.

Variables
Before propensity score matching After propensity score matching surgery, tumor size, and the presence of other tumors. The clinicopathological characteristics of patients with thymic epithelial tumor surgery before and after propensity score matching are presented in Table 2. Of the 2,451 patients with TETs, 348 received preoperative chemotherapy and 2,103 underwent direct surgery without preoperative chemotherapy. χ 2 tests showed significant differences in the proportion of patients  Table 3. Table 4 lists the 11 variables included in the univariate Cox regression model to analyze the factors associated with overall survival or cancer-specific survival in patients undergoing surgery for TETs. Variables with univariate analysis P < 0.1 were enrolled in multivariate Cox Regression models. Multivariate Cox regression analysis demonstrated that preoperative chemotherapy was an independent prognostic factor for OS (P = 0.002) and CSS (P = 0.013) in patients undergoing surgery for TETs. In addition, age at diagnosis, Masaoka-Koga staging, WHO classification, radiotherapy and lymph node biopsy findings were all independent prognostic factors for both OS and CSS.

Survival analysis and Forest plots
In the entire cohort before propensity score matching, Figures 2, 3 illustrate the prognosis of patients with TETs with different Masaoka-Koga staging who received preoperative chemotherapy vs. those who did not. Preoperative chemotherapy Survival curves of OS in the POCT group and no-POCT group in patients undergoing surgery for TETs at different stages before propensity matching.  Survival curves of CSS in the POCT group and no-POCT group in patients undergoing surgery for TETs at different stages before propensity matching.

FIGURE 4
Forest plots of OS and CSS in the POCT and no POCT groups in patients undergoing surgery for TETs at different stages before propensity matching.  Figure 4, with preoperative chemotherapy being a favorable factor for OS in patients with III/IV. In the propensity score-matched cohort, the overall survival and cancer-specific survival curves for the preoperative chemotherapy and no-preoperative chemotherapy groups are shown in  Survival curves of OS in the POCT group and no-POCT group in patients undergoing surgery for TETs after propensity matching.

Discussion
In this population-based study, we use data from the SEER database to evaluate the survival outcomes of 2,451 patients with TETs over the past 10-plus years. Compared to patients without preoperative chemotherapy, preoperative chemotherapy significantly improved OS in patients with stage III/IV TETs and did not significantly improve OS and CSS in stage I/IIA or IIB patients with TETs, both before and after propensity score matching. As a systemic treatment with some toxicities, preoperative chemotherapy significantly improved OS and CSS in younger patients or patients with multiple cancers with TETs. Therefore, we prefer to apply preoperative chemotherapy to FIGURE 9 Survival curves of OS in the POCT and no POCT groups in patients undergoing surgery for TETs in different age groups after propensity matching.
Fan et al. 10.3389/fsurg.2023.1108699 Frontiers in Surgery advanced TETs detected by imaging, and patient history and physical condition should be carefully considered when applying chemotherapy.
Whether chemotherapy can improve the prognosis of patients undergoing surgery for TETs has been controversial. A multicenter analysis in Japan reveals that chemotherapy did not provide any survival advantage for patients with completely resected stage III and IV thymoma and thymic carcinoma (19), but chemotherapy in the study was limited to postoperative chemotherapy. Wei et al. reported a higher 5year OS rate in patients with thymoma or thymic carcinoma who underwent direct surgery compared with those who received preoperative chemotherapy (20). In contrast, a study by Lucchi M et al. noted that patients who underwent surgery after neoadjuvant chemotherapy had better OS compared with patients who underwent primary surgery (21). Studies by FIGURE 10 Survival curves of CSS in the POCT and no POCT groups in patients undergoing surgery for TETs in different age groups after propensity matching.  (22,23). In our study population, the median tumor size of 94.01 mm in patients who opted for preoperative chemotherapy was greater than the median of 66.19 mm in the group without preoperative chemotherapy, implying that preoperative chemotherapy was more frequently administered in patients with larger tumors. Early case reports and some small review studies have also shown that chemo can help to reduce tumor size and relieve symptoms (24). In the study by Federico Venuta et al. it was also suggested that preoperative chemotherapy had a down-staging effect (22). In the study by Macchiarini et al. in which preoperative chemotherapy was administered to seven patients with invasive thymoma, all patients had at least a 50% reduction in tumor size (23). Almost all research has concluded that R0 resection is a determinant factor associated with thymic tumor survival FIGURE 11 Survival curves of OS in the POCT and no POCT groups in patients undergoing surgery for TETs in different wHO classification groups after propensity matching.  (28).
As first-line neoadjuvant therapy, the most popular chemotherapy regimens are platinum derivatives, mainly cisplatin with anthracyclines and/or etoposide, and they show good activity against both thymoma and thymic carcinoma, often with response rates above 50% (29). Carboplatin-paclitaxel is mainly recommended for thymic carcinoma. The main side effects FIGURE 12 Survival curves of CSS in the POCT and no POCT groups in patients undergoing surgery for TETs in different WHO classification groups after propensity matching.
Fan et al. 10.3389/fsurg.2023.1108699 reported for chemotherapy tend to be nausea and vomiting, bone marrow suppression, and cardiotoxicity of anthracyclines (28,30). Elderly patients have more complications and poor physical condition (10), which might decrease the tolerance of chemotherapy, so whether and when to use chemotherapy should be carefully chosen based on the patient's medical condition. A retrospective study by Samina Park et al. (31) suggested that the surgical group after neoadjuvant chemotherapy showed significantly higher transfusion rates (P = 0.003) and longer operative times (P < 0.001), but there was no evidence that neoadjuvant chemotherapy reduced long-term survival in patients with thymic epithelial tumors. The study by Cameron D et al. also concluded that chemotherapy induction followed by surgical treatment followed by radiotherapy is safe and probably the best sequence of treatment for carefully screened patients with advanced thymoma (32).

FIGURE 13
Survival curves of OS in the POCT and no POCT groups in patients with or without other malignancies undergoing surgery for TETs after propensity matching.
Fan et al. 10.3389/fsurg.2023.1108699 In addition, in our study, although the proportion of thymic carcinomas was smaller compared to thymomas, it appeared that thymic carcinomas responded more to induction chemotherapy, which is consistent with the findings of Robert et al. (27). Preoperative chemotherapy had several advantages over postoperative chemotherapy. For example, to prevent local tumor spread during surgery, to reduce tumor staging and thus improve surgical resection rates (33). Compared to postoperative chemotherapy, preoperative chemotherapy is better tolerated and most patients are able to reach surgery in good health, which also helps to improve the clinical status of the patient and to relieve symptoms (including myasthenia gravis remission) (24).
Previously, several SEER-based studies investigated the prognostic value of different treatment regimens for patients with TETs. However, we differ from previous studies by (1) using the SEER database for the first time to investigate the efficacy of preoperative chemotherapy in patients with thymic epithelial FIGURE 14 Survival curves of CSS in the POCT and no POCT groups in patients with or without other malignancies undergoing surgery for TETs after propensity matching.
Fan et al. 10.3389/fsurg.2023.1108699 Frontiers in Surgery tumors, (2) performing propensity score matching to increase comparability and reduce bias at baseline, and (3) performing subgroup analysis by factors such as Masaoka-Koga staging to find the characteristics of patients with TETs suitable for preoperative chemotherapy. The present study, like other SEER-based studies, has several limitations. First, although there is a wealth of data in the SEER database, it is not comprehensive and it lacks information on several important demographics, clinically relevant variables, and treatment modalities. For example, details of preoperative chemotherapy (including total dose of chemotherapy, daily fractions, and type of chemotherapeutic agent), whether the surgical margins were positive, and the patient's medical history and comorbidities. The lack of these variables leads to an incomplete clinical picture and potential bias, which may limit our assessment of preoperative chemotherapy. For this reason, we used the available data to focus on the impact of preoperative chemotherapy on long-term survival (≥1 month) in patients with thymic epithelial tumors. Second, although systemic treatment variables report the sequence of surgery and chemotherapy, they do not take into account the timing of events. Thus, it is possible that chemotherapy was administered more than 6 months prior to surgery or only 6 days prior to surgery. Chemotherapy administered prior to surgery may not have had sufficient time to anticipate the associated tumor response, which may underestimate the effect of preoperative chemotherapy. Although the SEER database is a representative national cancer registry with outstanding reliability and reproducibility of data collection and reporting procedures, and we used propensity score matching to minimize selection bias in preoperative chemotherapy, we could not completely exclude unmeasured or unpredictable confounding factors.

Conclusions
This study found that preoperative chemotherapy is a viable option for advanced thymoma with favorable overall and cancer-specific survival rates, but patient history and physical condition should be fully considered in conjunction with diagnostic imaging findings to assess patient tolerance to chemotherapy. Patients with thymic or multiple cancers may benefit from preoperative chemotherapy.

Data availability statement
The original contributions presented in the study are included in the article/Supplementary Materials, further inquiries can be directed to the corresponding author.