Systemic therapy for advanced thymic epithelial tumors: a narrative review of current evidence and perspectives Authors: Mari Tone, Takayuki Shiroyama Published: 2026 ABSTRACT Background and ObjectiveThymic epithelial tumors (TETs), including thymoma and thymic carcinoma, are rare tumors. Although systemic therapy is commonly used for unresectable advanced or recurrent TETs, establishing robust evidence remains challenging. This narrative review summarizes current evidence and incorporates the latest clinical trial data, reporting outcomes separately for thymoma and thymic carcinoma whenever possible. We also highlight practical considerations relevant to clinical practice, including dosing strategies and toxicity management.MethodsWe searched PubMed and Google Scholar for prospective and retrospective original studies, meta-analyses, review articles, and case studies published up to November 2025 using terms including “thymoma”, “thymic carcinoma”, “thymic cancer”, “thymic tumors”, “thymic epithelial tumors”, “chemotherapy”, “systemic treatment”, “immunotherapy”, and “targeted therapy”. Ongoing trials were identified via ClinicalTrials.gov and the Japan Registry of Clinical Trials. Retrospective studies enrolling fewer than 15 participants were excluded.Key Content and FindingsSystemic therapy for advanced TETs continues to include cytotoxic chemotherapy as a major component. In selected settings, particularly in thymic carcinoma, targeted therapy and immune checkpoint inhibitors (ICIs) represent additional active options. Emerging trials suggest that combination strategies may enhance efficacy. Novel drugs, including bispecific antibodies and antibody–drug conjugates, are being evaluated in ongoing phase II trials.ConclusionsDespite the rarity of TETs, systemic treatment options have gradually expanded through the accumulation of evidence from small phase II studies. Future advances will require collaborative prospective trials to validate predictive biomarkers and support biomarker-enriched trial designs. Introduction Thymic epithelial tumors (TETs) are a heterogeneous group of rare tumors arising from thymic epithelium, which plays an important role in the maturation of T lymphocytes. Thymomas typically retain organotypic features of the normal thymus and often follow a relatively indolent clinical course, whereas thymic carcinomas lack these features and instead exhibit overt cytologic atypia with more aggressive behavior (1). According to the 5th edition of the World Health Organization (WHO) classification of thoracic tumors (1), tumors of the thymus are broadly categorized into thymoma, thymic carcinoma, and thymic neuroendocrine neoplasms. In this article, we use the term TETs to refer specifically to thymoma and thymic carcinoma and do not address thymic neuroendocrine neoplasms. At initial presentation, the stage distribution differs between thymoma and thymic carcinoma. Although most patients with thymoma present with early-stage disease, a subset are diagnosed with stage III or IV disease (2). In contrast, thymic carcinoma is more often detected at an advanced or metastatic stage, with more than three-quarters of patients presenting with stage III or IV disease (2). Prognosis varies substantially according to stage and histologic subtype. In large retrospective studies, the 5-year survival rates for thymoma were approximately 70–90% for Masaoka stage III and 40–60% for stage IV, whereas those for thymic carcinoma were approximately 50–60% and 20–40% for stage III and IV, respectively (3-6). These data highlight that survival remains unsatisfactory in patients with advanced-stage disease, particularly those with thymic carcinoma, and they emphasize the need to optimize systemic treatment strategies for advanced TETs. Although systemic therapy is commonly used for unresectable advanced or recurrent TETs, it is difficult to obtain sufficient evidence of systemic therapy for TETs because TETs are rare tumors. The current treatment strategy is formulated largely based on small-scale phase II trials or retrospective cohort studies. Cytotoxic anticancer agents are still the mainstream of systemic therapy for TETs; molecular-targeted drugs and immune checkpoint inhibitors (ICIs) have also demonstrated a certain degree of efficacy for TETs. Furthermore, recent clinical trials have shown that combining these drugs can produce better therapeutic effects. In this narrative review, we summarize the current evidence on systemic therapies for advanced TETs, incorporating the most recent clinical trial data. Given that thymoma and thymic carcinoma differ in key clinicopathologic features yet are frequently pooled in clinical trials, we report outcomes separately by histologic type (thymoma vs. thymic carcinoma) where data are available. We also discuss practical considerations for treatment delivery and toxicity management in routine practice and outline future directions for advanced TETs. We present this article in accordance with the Narrative Review reporting checklist (available at https://med.amegroups.com/article/view/10.21037/med-2025-1-77/rc). Methods The current narrative review included prospective and retrospective original studies, meta-analyses, review articles, and case studies written in English and searched via PubMed and Google Scholar, considering all articles published up to November 2025. Ongoing and unpublished trials were additionally identified through ClinicalTrials.gov and the Japan Registry of Clinical Trials (Table 1). The terms used for literature search were “thymoma”, “thymic carcinoma”, “thymic cancer”, “thymic tumors”, “thymic epithelial tumors”, “chemotherapy”, “systemic treatment”, “immunotherapy”, and “targeted therapy”. The studies on neoadjuvant and postoperative adjuvant chemotherapy or concurrent combined chemoradiation therapy were excluded. We also excluded retrospective studies with fewer than 15 participants because small sample sizes limit the reliability and interpretability of outcome estimates. When trials enrolled both thymoma and thymic carcinoma, we extracted and reported treatment outcomes separately by histologic subtype where available. Therefore, sample sizes shown in the thymoma- and thymic carcinoma-specific sections and tables represent histology-specific subsets when studies enrolled mixed histologies, rather than the total trial accrual. Table 1The search strategy summaryItemsSpecificationDate of searchSeptember 1, 2025 to November 20, 2025Databases and other sources searchedPubMed, Google Scholar, ClinicalTrials.gov, Japan Registry of Clinical TrialsSearch terms“Thymoma”, “thymic carcinoma”, “thymic cancer”, “thymic tumors”, “thymic epithelial tumors”, “chemotherapy”, “systemic treatment”, “immunotherapy”, and “targeted therapy”TimeframeUp to November 2025Inclusion criteriaOriginal articles, meta-analyses, review articles, and case studies written in EnglishExclusion criteriaThe studies on neoadjuvant and postoperative adjuvant chemotherapy or concurrent combined chemoradiation therapy were excluded. We also excluded retrospective studies with fewer than 15 participantsSelection processBoth authors searched the literature and subsequently discussed and agreed on the literature The search strategy summary First-line systemic therapy for advanced thymoma Several prospective phase II trials or retrospective cohort studies have reported first-line systemic therapy for stage III or IV advanced thymoma (Table 2, Figure 1). Although trials designed to deliver perioperative (neoadjuvant or adjuvant) chemotherapy or chemoradiation were excluded, surgery or radiotherapy was sometimes administered after first-line chemotherapy in advanced thymoma, which can confound time-to-event endpoints; progression-free survival (PFS) and overall survival (OS) should therefore be interpreted with caution. Table 2First-line systemic therapy for advanced thymomaRegimenYear (ref.)Study designNumberStageTypeORR (%)PFS (months)OS (months)Anthracycline ADOC1991 (7)Retrospective37III/IVA/B91.812.0a15.01999 (8)Retrospective16III/IVaA/B81.233.2b,c47.5c CAP1994 (9)Phase II30dIII/IVNA50.0d11.8a,d37.7d CODE2009 (10)Phase II27IVB59.09.573.2 CAP + belinostat2014 (11)Phase I/II11eIVB64.0NRNR Carboplatin + amrubicin2014 (12)Phase II15eNAA/B13.37.6f58.0fNon-anthracycline PE1996 (13)Phase II16III/IVB56.026.451.6 VIP2001 (14)Phase II20eIII/IVNA35.011.9a,g31.6g2010 (15)Phase II16hIII/IVAB/B25.0h13.1a,hNRh CP2011 (16)Phase II21eIII/IVA/B42.916.7NRPreferred first-line regimen per the NCCN guidelines (v2.2025) is CAP. a, median duration of response. b, disease-free interval. c, includes cases resected after chemotherapy. d, 1 TC patient was included. e, while the clinical trial included both T and TC patients, this table presented data only for T patients. f, combined data from untreated and previously treated patients. g, 8 TC patients were included. h, 4 TC patients were included. ADOC, cisplatin + doxorubicin + vincristine + cyclophosphamide; CAP, cisplatin + doxorubicin + cyclophosphamide; CODE, cisplatin + vincristine + doxorubicin + etoposide; CP, carboplatin + paclitaxel; NA, not available; NCCN, National Comprehensive Cancer Network; NR, not reached; ORR, overall response rate; OS, overall survival; PE, cisplatin + etoposide; PFS, progression-free survival; ref., reference; T, thymoma; TC, thymic carcinoma; VIP, etoposide + ifosfamide + cisplatin. First-line systemic therapy for advanced thymoma Preferred first-line regimen per the NCCN guidelines (v2.2025) is CAP. a, median duration of response. b, disease-free interval. c, includes cases resected after chemotherapy. d, 1 TC patient was included. e, while the clinical trial included both T and TC patients, this table presented data only for T patients. f, combined data from untreated and previously treated patients. g, 8 TC patients were included. h, 4 TC patients were included. ADOC, cisplatin + doxorubicin + vincristine + cyclophosphamide; CAP, cisplatin + doxorubicin + cyclophosphamide; CODE, cisplatin + vincristine + doxorubicin + etoposide; CP, carboplatin + paclitaxel; NA, not available; NCCN, National Comprehensive Cancer Network; NR, not reached; ORR, overall response rate; OS, overall survival; PE, cisplatin + etoposide; PFS, progression-free survival; ref., reference; T, thymoma; TC, thymic carcinoma; VIP, etoposide + ifosfamide + cisplatin. Therapeutic efficacy of TETs. Results of selected clinical studies are shown by treatment line for thymoma and thymic carcinoma. The bubble size is proportional to the sample size of each study. The figure is descriptive and not intended for indirect comparisons across trials. It displays a selected subset of studies from Tables 2-5; studies with mixed treatment lines or mixed histologies without separable data, and very small cohorts were not plotted. ADOC, cisplatin + doxorubicin + vincristine + cyclophosphamide; CAP, cisplatin + doxorubicin + cyclophosphamide; CODE, cisplatin + vincristine + doxorubicin + etoposide; CP, carboplatin + paclitaxel; ICI, immune checkpoint inhibitor; ORR, overall response rate; PD-1, programmed death-1; PE, cisplatin + etoposide; PFS, progression-free survival; TET, thymic epithelial tumor; VIP, etoposide + ifosfamide + cisplatin. Given the lack of head-to-head comparative trials and heterogeneity among study populations and designs, the available evidence is summarized descriptively and is not intended for indirect cross-trial comparisons or superiority claims. Anthracycline-based regimens In the first-line treatment of thymoma, anthracycline-based regimens are among the earliest reported and remain the mainstay of contemporary clinical practice. In retrospective series, the ADOC regimen (cisplatin + doxorubicin + vincristine + cyclophosphamide) achieved a notably high overall response rate (ORR) of 81–92% (7,8). Given its four-drug composition, the ADOC regimen necessitates careful toxicity management, including cardiotoxicity, gastrointestinal toxicity, and vincristine-associated peripheral neuropathy. The CAP regimen (cisplatin + doxorubicin + cyclophosphamide) omits vincristine from ADOC and may therefore reduce the risk of neuropathy. In the phase II study of CAP, a comparatively favorable efficacy profile was reported, with an ORR of 50%, a median duration of response of 11.8 months, and a median OS of 37.7 months (9). When interpreting the results of this CAP trial, it is important to note that many patients underwent resection after chemotherapy. The CODE regimen (cisplatin + vincristine + doxorubicin + etoposide) showed antitumor activity in disseminated thymoma, achieving an ORR of 59%. The reported median OS was over 6 years, which the investigators suggested may have been influenced by subsequent local therapy, particularly in stage IVa disease (10). Non-anthracycline-based regimens Several non-anthracycline cisplatin-based regimens, such as PE (cisplatin + etoposide) and VIP (etoposide + ifosfamide + cisplatin), also showed activity in small phase II studies, with reported ORRs of 25–56% (13-15). Because these regimens omit anthracycline, they can avoid anthracycline-associated cardiotoxicity and may therefore be considered for patients with pre-existing cardiac disease or in older adults. For patients who are not candidates for cisplatin, carboplatin-paclitaxel has also been investigated as an alternative, with an ORR of 42.9% and a median PFS of 16.7 months (16). Treatment selection for treatment-naïve thymoma To date, no practice-changing prospective evidence has emerged to establish a new standard first-line systemic therapy for treatment-naïve advanced thymoma. Accordingly, the optimal first-line regimen remains uncertain. A pooled analysis of platinum-based chemotherapy for advanced thymoma and thymic carcinoma including 15 clinical trials and retrospective data reported a significantly higher ORR for platinum plus anthracycline-based chemotherapy than for platinum plus non-anthracycline-based chemotherapy in thymoma (69.4% vs. 37.8%, P<0.0001) (17). Although subject to interstudy heterogeneity and potential selection bias, these data may support platinum-anthracycline combinations as a reasonable default for first-line treatment of advanced thymoma. Consistent with this, the latest National Comprehensive Cancer Network (NCCN) guidelines list the CAP regimen as a preferred regimen, with other anthracycline- or platinum-based combinations (e.g., ADOC; PE/VIP in selected contexts) noted as reasonable alternatives (2). Given the frequency of thymoma-associated paraneoplastic autoimmune syndromes (e.g., myasthenia gravis, pure red cell aplasia, and Good’s syndrome), baseline clinical evaluation, including neuromuscular autoantibody testing such as anti-acetylcholine receptor (AChR) antibodies and assessments of reticulocyte counts and immunoglobulin levels, should be incorporated into treatment planning and monitoring when initiating systemic therapy. Paraneoplastic autoimmune syndromes may influence regimen selection and monitoring intensity, and therefore warrant individualized, multidisciplinary treatment planning. Among the commonly used platinum-based regimens, the main severe toxicities are hematologic (notably neutropenia/febrile neutropenia) and gastrointestinal adverse events, whereas anthracycline-containing regimens also warrant attention regarding cardiac toxicity. When toxicity is acceptable, a cisplatin–anthracycline regimen is considered first; however, its administration requires vigilant toxicity monitoring and proactive supportive care. In patients with pre-existing cardiac disease or in older patients, it is prudent to consider substitution with a non-anthracycline regimen or a carboplatin-based combination to mitigate treatment-related risk. Second- and subsequent-line systemic therapy for advanced thymoma In second- and subsequent-line treatments for advanced thymoma, prospective evidence remains limited (Table 3, Figure 1). Although first-line regimens for advanced thymoma predominantly consist of cytotoxic chemotherapy, second- and subsequent-line trials are more diverse and include not only cytotoxic agents but also targeted agents and ICIs, often in trials that pooled thymoma participants with thymic carcinoma participants. Table 3Second- and subsequent-line systemic therapy for advanced thymomaRegimenYear (ref.)Study designNumberStageTypeORR (%)PFS (months)OS (months)Cytotoxic agents Carboplatin + amrubicin2014 (12)Phase II3aNAA/B33.37.6b58.0b Capecitabine + gemcitabine2014 (18)Phase II22aIVB40.0c11.0NA Pemetrexed2018 (19)Phase II16aIVB25.012.146.4 Amrubicin2019 (20)Phase II14aIVB2/329.08.354.1 S-12020 (21)Phase II20aIII/IVAB/B10.011.358.5 Ifosfamide2023 (22)Prospective5aIVB60.013.2NATargeted agents Sunitinib2015 (23)Phase II16aNAB6.08.515.52023 (24)Phase II12aIVB307.747.9 Everolimus2018 (25)Phase II32aIII/IVA/AB/B9.416.6NR Bevacizumab + CPd2022 (26)Retrospective21aNAB2/323.86.0NA Palbociclib2023 (27)Phase II24aIVA/B16.713.026.4 Cixutumumab2014 (28)Phase II37aNAAB/B13.59.927.5 Selinexor2025 (29)Phase II16aIII/IVB12.513.6NRICIs Pembrolizumab2019 (30)Phase II7aIVB2/328.66.1NR Pembrolizumab + lenvatinib2025 (31)Phase II7aIVB328.6NRNA Avelumab2019 (32)Phase I7aNAB28.6NANA Avelumab + axitinib2022 (33)Phase II5aIVB340.07.5e26.6e Atezolizumab2022 (34)Phase II13fIII/IVNA38.511.8NA Nivolumab2023 (35)Phase II53gNAB314.3g6.2g21.3gOthers Octreotide + prednisone2002 (36)Phase II10aII–IVB40.014h,i15h Octreotied ± prednisone2004 (37)Phase II32aIII/IVNA37.58.8NRPreferred second-line options per the NCCN guidelines (v2.2025) include gemcitabine ± capecitabine, pemetrexed, everolimus, and octreotide (if octreotide-avid) ± prednisone. a, while the clinical trial included both T and TC patients, this table presented data only for T patients. b, combined data from untreated and previously treated patients. c, 8 TC patients were included. d, platinum + paclitaxel. e, 27 TC patients were included. f, while the clinical trial included several kinds of solid-tumor patients, this table presented data only for T patients. g, 10 TC patients were included. h, 6 TC patients were included. i, median time to progression. CP, carboplatin + paclitaxel; ICI, immune checkpoint inhibitor; NA, not available; NCCN, National Comprehensive Cancer Network; NR, not reached; ORR, overall response rate; OS, overall survival; PFS, progression-free survival; ref., reference; T, thymoma; TC, thymic carcinoma. Second- and subsequent-line systemic therapy for advanced thymoma Preferred second-line options per the NCCN guidelines (v2.2025) include gemcitabine ± capecitabine, pemetrexed, everolimus, and octreotide (if octreotide-avid) ± prednisone. a, while the clinical trial included both T and TC patients, this table presented data only for T patients. b, combined data from untreated and previously treated patients. c, 8 TC patients were included. d, platinum + paclitaxel. e, 27 TC patients were included. f, while the clinical trial included several kinds of solid-tumor patients, this table presented data only for T patients. g, 10 TC patients were included. h, 6 TC patients were included. i, median time to progression. CP, carboplatin + paclitaxel; ICI, immune checkpoint inhibitor; NA, not available; NCCN, National Comprehensive Cancer Network; NR, not reached; ORR, overall response rate; OS, overall survival; PFS, progression-free survival; ref., reference; T, thymoma; TC, thymic carcinoma. The available evidence is summarized descriptively and is not intended for indirect cross-trial comparisons or superiority claims. Cytotoxic agents for pretreated thymoma Cytotoxic agents remain the mainstream for second- and subsequent-line treatments for advanced thymoma. However, the available evidence is limited to small, single-arm studies. Among single agents, pemetrexed and amrubicin showed modest response rates with median PFS of 8–12 months (19,20). Oral S-1 yielded a low ORR but a comparable median PFS, likely reflecting a high disease control rate (95%, 19 of 20) (21). Combination cytotoxic regimens have been evaluated in small phase II cohorts, with ORRs ranging from 33% to 40% and a median PFS of approximately 8–11 months across studies (12,18). There is also a small amount of prospective data available for ifosfamide (n=5), but the limited sample size leads to significant uncertainty (22). Targeted agents, ICIs, and other agents for pretreated thymoma Compared with thymic carcinoma, current evidence suggests that targeted agents have generally yielded low ORRs in thymoma. In multiple phase II studies of single-agent targeted therapies, ORRs ranged from 0% to 17% (23-25,27-29). Despite modest ORRs, median PFS exceeded one year with selected agents, including everolimus, palbociclib, and selinexor (25,27,29). In a single-center retrospective study, platinum-paclitaxel plus bevacizumab showed activity in previously treated thymoma, with an ORR of approximately 24% (26). Although ICIs have demonstrated antitumor activity for patients with previously treated thymoma (30-35), thymoma-specific evidence remains limited to very small patient subsets extracted from small trials enrolling mixed tumor types. Their clinical use is substantially limited by the high frequency of severe immune-related adverse events (irAEs), as detailed later (see the “Practical strategies for treatment delivery and toxicity management” section). Treatment selection for pretreated thymoma Determining the optimal second- or later-line systemic therapy for thymoma is challenging because the clinical efficacy of available regimens is generally limited. Targeted agents appear to show limited antitumor activity in thymoma relative to thymic carcinoma, and ICIs should generally be avoided outside clinical trials because of the high incidence of severe irAEs in thymoma. Consequently, cytotoxic chemotherapy remains the preferred default beyond the first-line therapy, with regimen selection individualized according to prior treatments and patient tolerability. First-line systemic therapy for advanced thymic carcinoma Cytotoxic chemotherapy remains a key component of first-line treatment for advanced thymic carcinoma, as in advanced thymoma. However, emerging prospective phase II evidence suggests that combination strategies incorporating anti-angiogenic agents or ICIs may improve outcomes in advanced thymic carcinoma (e.g., the RELEVENT and MARBLE trials), marking a shift in the systemic therapy landscape (Table 4, Figure 1). Table 4First-line systemic therapy for advanced thymic carcinomaRegimenYear (ref.)Study designNumberStageORR (%)PFS (months)OS (months)Cytotoxic agents VIP2001 (14)Phase II8aIII/IV25.011.9b,c31.6b2010 (15)Phase II4aIII/IV25.013.1c,dNRd ADOC2011 (38)Retrospective34IV50.0NA21.3 CP2011 (16)Phase II23aIII/IV21.75.020.02011 (39)Retrospective16IV37.58.649.42015 (40)Phase II39III/IV35.97.5NR2016 (41)Prospective37IV29.76.043.0 Carboplatin + amrubicin2014 (12)Phase II19aNA42.17.6e27.3e CAP + belinostat2014 (11)Phase I/II14aIV21.07.221.4 Cisplatin + irinotecan2021 (42)Retrospective18IV44.48.445.6Targeted agents Ramucirumab + CP2024 (43)Phase II35IV57.618.143.82024 (44)Randomized phase II8fIV88.08.0NA CP (control arm)2024 (44)Randomized phase II12fIV40.07.0NAICIs CP + atezolizumab2025 (45)Phase II48III/IV56.09.6NR TP + PD-1 inhibitor2025 (46)Retrospective47IV44.79.452.3Preferred first-line regimens per the NCCN guidelines (v2.2025) include CP and CP + ramucirumab. a, while the clinical trial included both T and TC patients, this table presented data only for TC patients. b, 20 T patients were included. c, median duration of response. d, 15 T patients were included. e, combined data from untreated and previously treated patients. f, terminated early owing to slow accrual. ADOC, cisplatin + doxorubicin + vincristine + cyclophosphamide; CAP, cisplatin + doxorubicin + cyclophosphamide; CP, carboplatin + paclitaxel; ICI, immune checkpoint inhibitor; NA, not available; NCCN, National Comprehensive Cancer Network; NR, not reached; ORR, overall response rate; OS, overall survival; PD-1, programmed death-1; PFS, progression-free survival; ref., reference; T, thymoma; TC, thymic carcinoma; TP, taxane + platinum; VIP, etoposide + ifosfamide + cisplatin. First-line systemic therapy for advanced thymic carcinoma Preferred first-line regimens per the NCCN guidelines (v2.2025) include CP and CP + ramucirumab. a, while the clinical trial included both T and TC patients, this table presented data only for TC patients. b, 20 T patients were included. c, median duration of response. d, 15 T patients were included. e, combined data from untreated and previously treated patients. f, terminated early owing to slow accrual. ADOC, cisplatin + doxorubicin + vincristine + cyclophosphamide; CAP, cisplatin + doxorubicin + cyclophosphamide; CP, carboplatin + paclitaxel; ICI, immune checkpoint inhibitor; NA, not available; NCCN, National Comprehensive Cancer Network; NR, not reached; ORR, overall response rate; OS, overall survival; PD-1, programmed death-1; PFS, progression-free survival; ref., reference; T, thymoma; TC, thymic carcinoma; TP, taxane + platinum; VIP, etoposide + ifosfamide + cisplatin. The available evidence is summarized descriptively and is not intended for indirect cross-trial comparisons or superiority claims. Cytotoxic agents for treatment-naïve thymic carcinoma Evidence for first-line cytotoxic agents in advanced thymic carcinoma is largely limited to small cohorts, most commonly using platinum-based regimens. Carboplatin-paclitaxel is the most frequently reported platform, with ORRs of approximately 20–40% and a median PFS of 5–9 months across studies (16,39-41). Anthracycline-based cisplatin regimens have shown variable activity (ORR approximately 20–50%) in small cohorts (11,12,38), which appears less robust than the responses reported in thymoma (7-9). The VIP regimen was evaluated in a phase II trial that enrolled both thymoma and thymic carcinoma; thymic carcinoma-specific outcomes were available only for a very limited subset (n=4–8) (14,15). Carboplatin-paclitaxel may offer a more favorable safety profile, particularly regarding cardiac toxicity, while potentially maintaining an efficacy comparable to that of anthracycline-based regimens. As detailed below, recent first-line development has increasingly built on the carboplatin-paclitaxel regimen, incorporating molecular-targeted agents or ICIs in combination strategies. Targeted agents for treatment-naïve thymic carcinoma While data on first-line targeted therapy in advanced thymic carcinoma remain limited, recent clinical development has increasingly focused on vascular endothelial growth factor receptor 2 (VEGFR2) blockade with ramucirumab. In the prospective, single-arm phase II RELEVENT trial (n=35), ramucirumab plus carboplatin-paclitaxel has shown promising activity, with an ORR of 57.6%, a median PFS of 18.1 months, and a median OS of 43.8 months (43). In SWOG S1701, a small randomized phase II trial that closed early because of slow accrual (n=20), the addition of ramucirumab to carboplatin-paclitaxel increased the ORR (88% vs. 40%, P=0.04), whereas PFS did not differ significantly [hazard ratio (HR) =0.51; P=0.13] (44); the interpretation is limited by the small sample size. Anti-angiogenic agents were generally manageable but required careful patient selection and vigilant monitoring for class-specific toxicities. In the RELEVENT trial, ramucirumab-related serious events included four cardiovascular events: two grade 3 acute myocardial infarctions, one grade 4 pulmonary embolism, and one grade 3 arterial hemorrhage (43). These safety signals warrant careful cardiovascular risk assessment, blood-pressure control, and surveillance for thromboembolism and bleeding during anti-VEGFR2 therapy. ICIs for treatment-naïve advanced thymic carcinoma As in other malignancies, ICI-chemotherapy combinations have been explored in advanced thymic carcinoma. In the prospective, multicenter, single-arm phase II MARBLE study (n=48), first-line carboplatin-paclitaxel plus atezolizumab achieved an ORR of 56.0% and a median PFS of 9.6 months; median OS was not reached at a median follow-up of 15.3 months (45). These prospective data support chemoimmunotherapy as an emerging first-line option for thymic carcinoma. In a separate real-world retrospective, non-randomized cohort study from China (n=107), platinum-taxane plus programmed death-1 (PD-1) inhibitor was associated with longer PFS than platinum-taxane alone (9.4 vs. 6.3 months; HR =0.37; P<0.001) (46). However, these findings should be interpreted cautiously given the retrospective, non-randomized design. Treatment selection for treatment-naïve thymic carcinoma In the absence of prospective comparative trials, the optimal first-line regimen for advanced thymic carcinoma has not been definitively established. In the pooled analysis of platinum-based chemotherapy across TETs, the ORR in thymic carcinoma was similar between platinum-anthracycline and platinum-non-anthracycline regimens (41.8% vs. 40.9%, P=0.91), in contrast to the superiority observed in thymoma (17). In the latest NCCN guidelines, anthracycline-containing regimens are not listed as preferred first-line options for thymic carcinoma but remain included as “Other Recommended” regimens, in part due to toxicity concerns (2). Among preferred first-line options, reflecting emerging phase II evidence, the guidelines list carboplatin-paclitaxel and carboplatin-paclitaxel plus ramucirumab as preferred options for treatment-naïve advanced thymic carcinoma (2,43). Chemoimmunotherapy combinations are expected to play an increasingly central role in first-line management of advanced thymic carcinoma. However, the combination therapy requires vigilant toxicity monitoring and proactive management, including baseline risk assessment, timely dose modifications, and coordinated supportive care to mitigate overlapping adverse events. Second- and subsequent-line systemic therapy for advanced thymic carcinoma For second- and later-line treatment for advanced thymic carcinoma, various agents, including targeted agents and ICIs, have been evaluated in recent years (Table 5, Figure 1). Table 5Second- and subsequent-line systemic therapy for advanced thymic carcinomaRegimenYear (ref.)Study designNumberStageORR (%)PFS (months)OS (months)Cytotoxic agents Carboplatin + amrubicin2014 (12)Phase II14aNA14.37.6b27.3b Capecitabine + gemcitabine2014 (18)Phase II8aNA40.0c6.0NA S-12020 (21)Phase II20aIII/IV25.05.422.72020 (47)Phase II26IV30.84.327.4 Pemetrexed2018 (19)Phase II11aIV9.12.99.8 Amrubicin2019 (20)Phase II19aIV11.07.318.0 Ifosfamide2023 (22)Prospective13aIV15.03.510.5Targeted agents Sunitinib2015 (23)Phase II23aNA26.07.2NR2016 (48)Retrospective20aIII/IV20.03.312.32023 (24)Phase II31aIII/IV21.48.827.8 Everolimus2018 (25)Phase II18aIII/IV16.75.614.7 Lenvatinib2020 and 2024 (49,50)Phase II42III/IV38.19.328.32025 (51)Retrospective22IV27.37.8NR2025 (52)Retrospective87III/IV30.010.2NR Bevacizumab + CPd2022 (26)Retrospective28aNA57.18.0NA Palbociclib2023 (27)Phase II23aIV8.79.225.6 Cixutumumab2014 (28)Phase II12aNA01.78.4 Selinexor2025 (29)Phase II15III/IV6.77.815.5ICI monotherapy Pembrolizumab2018 (53)Phase II40III/IV22.54.224.92019 (30)Phase II26aIV19.26.114.5 Nivolumab2019 (54)Phase II15eIII/IV03.814.12023 (35)Phase II53fNA14.3f6.2f21.3f PD-1 inhibitor2022 (55)Retrospective23III/IV17.42.126.7 Atezolizumab2025 (56)Phase II34IV14.73.2NRTargeted agents + ICIs Axitinib + avelumab2022 (33)Phase II27aIV33.07.5g26.6g Vorolanib + nivolumab2024 (57)Phase I/II9hNA11.19.121.1 Lenvatinib + pembrolizumab2025 (31)Phase II36aIV22.211.1NAOthers Octreotide ± prednisone2004 (37)Phase II6aIII/IV04.523.4Preferred second-line options per the NCCN guidelines (v2.2025) include gemcitabine ± capecitabine, sunitinib, lenvatinib, and pembrolizumab. a, while the clinical trial included both T and TC patients, this table presented data only for TC patients. b, combined data from untreated and previously treated patients. c, 22 T patients were included. d, platinum + paclitaxel. e, terminated early per the prespecified futility criterion. f, 43 T patients were included. g, 5 T patients were included. h, while the clinical trial included several kinds of solid-tumor patients, this table presented data only for TC patients. CP, carboplatin + paclitaxel; ICI, immune checkpoint inhibitor; NA, not available; NCCN, National Comprehensive Cancer Network; NR, not reached; ORR, overall response rate; OS, overall survival; PD-1, programmed death-1; PFS, progression-free survival; ref., reference; T, thymoma; TC, thymic carcinoma. Second- and subsequent-line systemic therapy for advanced thymic carcinoma Preferred second-line options per the NCCN guidelines (v2.2025) include gemcitabine ± capecitabine, sunitinib, lenvatinib, and pembrolizumab. a, while the clinical trial included both T and TC patients, this table presented data only for TC patients. b, combined data from untreated and previously treated patients. c, 22 T patients were included. d, platinum + paclitaxel. e, terminated early per the prespecified futility criterion. f, 43 T patients were included. g, 5 T patients were included. h, while the clinical trial included several kinds of solid-tumor patients, this table presented data only for TC patients. CP, carboplatin + paclitaxel; ICI, immune checkpoint inhibitor; NA, not available; NCCN, National Comprehensive Cancer Network; NR, not reached; ORR, overall response rate; OS, overall survival; PD-1, programmed death-1; PFS, progression-free survival; ref., reference; T, thymoma; TC, thymic carcinoma. Alongside cytotoxic chemotherapy, selected targeted agents and ICIs are recognized as subsequent-line options. More recently, trials evaluating combinations of molecular-targeted therapies with ICIs have also emerged. Practice-informing evidence for second-line and subsequent therapy in advanced thymic carcinoma is primarily derived from prospective phase II studies, which include NCCN-preferred regimens such as lenvatinib (REMORA trial), sunitinib, and pembrolizumab. The available evidence is summarized descriptively and is not intended for indirect cross-trial comparisons or superiority claims. Cytotoxic agents for pretreated thymic carcinoma Prospective evidence for cytotoxic agents in previously treated thymic carcinoma is limited and is largely derived from small, single-arm studies (12,18-22,47). Single cytotoxic agents have generally shown modest activity (ORR 9–15%; median PFS 3–7 months) (19,20,22). Notably, oral S-1 has demonstrated comparatively favorable activity in Japanese phase II cohorts (ORR 25–30%; median OS 22–27 months) (21,47), although its availability is region dependent. Combination cytotoxic regimens have also been explored in small cohorts, but interpretation is constrained by sample size and clinical heterogeneity (12,18). Targeted agents for pretreated thymic carcinoma Several targeted agents have been evaluated in previously treated thymic carcinoma, with the most notable activity reported for anti-angiogenic multi-kinase inhibitors. In the multicenter, prospective phase II REMORA trial (n=42), lenvatinib achieved an ORR of 38% and a median PFS of 9.3 months (49), with long-term follow-up reporting a median OS of 28.3 months (50). Real-world retrospective studies are consistent with these findings and suggest clinical activity in routine practice (51,52). Sunitinib also showed antitumor activity in two phase II studies, with ORRs of 21–26% and a median PFS of approximately 7–9 months (23,24). Other targeted approaches, including IGF-1R blockade with cixutumumab and the XPO1 inhibitor selinexor, have shown limited activity (28,29). ICIs for pretreated thymic carcinoma ICIs have shown modest activity for pretreated thymic carcinoma in small, single-arm phase II studies. Pembrolizumab achieved ORRs of 19–23% with a median PFS of 4–6 months (30,53), whereas nivolumab yielded heterogeneous results (ORR 0% in PRIMER and 14% in NIVOTHYM trial) (35,54). Atezolizumab has also been evaluated, with modest activity in a phase II trial (56). Reported outcomes have varied across studies, but cross-trial comparisons are inherently limited and may be influenced by small sample sizes and differences in patient/tumor characteristics, including programmed death-ligand 1 (PD-L1) distribution. Regarding safety, although irAEs appear less frequent than in thymoma, myocarditis and other neuromuscular immune events remain key concerns with ICIs in thymic carcinoma. Combination with targeted agents and ICIs for pretreated thymic carcinoma Combination strategies integrating targeted agents with ICIs have also been evaluated in pretreated advanced thymic carcinoma. In the phase II PECATI study (31), lenvatinib plus pembrolizumab met its primary endpoint (5-month PFS rate of 88.4%) and achieved a median PFS of 11.1 months with an ORR of 22.2% in the thymic carcinoma subgroup. In the phase II CAVEATT trial (33), axitinib plus avelumab yielded an ORR of 33% with a median OS of 26.6 months. Early-phase evaluation of nivolumab plus the anti-angiogenic tyrosine kinase inhibitor (TKI) vorolanib has also been reported in refractory thoracic tumors, including thymic carcinoma, but evidence remains limited (57). Across these regimens, toxicity reflects both VEGFR-inhibitor effects and immune-related events, requiring heightened vigilance; dose modifications and supportive care are essential (see the “Practical strategies for treatment delivery and toxicity management” section). Treatment selection for pretreated thymic carcinoma Given the limited, mostly single-arm evidence base in pretreated thymic carcinoma, treatment selection should be individualized based on prior therapy, clinical status, and toxicity risk. More recent data support lenvatinib as one of the more active options in pretreated advanced disease, and ICIs have demonstrated clinical activity in selected patients. In line with these developments, the latest NCCN guidelines list lenvatinib and pembrolizumab among preferred agents for second- or later-line therapy (2). In the future, combination strategies that incorporate molecular-targeted agents may be poised to assume a more prominent role beyond the first-line. Practical strategies for treatment delivery and toxicity management Treat-to-tolerability strategies for lenvatinib Lenvatinib is a key targeted option for previously treated thymic carcinoma. In the REMORA phase II study (n=42), all patients had at least one dose reduction of lenvatinib due to adverse events, most commonly hypertension, proteinuria, and hand-foot syndrome. Although 9.5% of patients discontinued treatment due to adverse events, no treatment-related deaths were reported (49). A subsequent long-term follow-up and exploratory analysis showed that a higher relative dose intensity (RDI) at 8 weeks, defined as ≥75% RDI, was significantly associated with improved OS compared to RDI <75% (median OS, 38.5 vs. 17.3 months; HR =0.46; P=0.0406) (50). In addition, patients who maintained their lenvatinib dose during the first 8 weeks had a significantly higher ORR than those who required dose reduction (75.0% vs. 29.4%, P=0.0379) (50). These exploratory findings suggest that maintaining adequate early treatment intensity is associated with more favorable outcomes in this setting. This supports the need for careful monitoring and prompt management of adverse events during the initial weeks to preserve dose intensity. Nevertheless, even with careful management of adverse events, many patients require dose reduction or treatment interruptions because of cumulative toxicities. In this context, alternative dosing approaches aimed at preserving treatment exposure while reducing toxicity have been explored for other tumor types. In unresectable hepatocellular carcinoma, a retrospective study reported that weekends-off administration of lenvatinib (5-day-on/2-day-off administration) reduced toxicities and improved tolerability, contributed to a longer treatment duration and prolonged OS (58). Furthermore, in a post-hoc analysis of the prospective COLLECT study in thyroid cancer (59), planned drug holidays of lenvatinib were associated with significantly improved OS and other clinical outcomes compared with continuous daily administration. In that study, planned drug holidays were utilized as a strategy to avoid severe or intolerable toxicities. These findings suggest that thoughtfully implemented drug-holiday strategies may help mitigate toxicity while preserving clinically meaningful dose intensity. However, the supporting evidence remains limited and derives from other tumor types. Data on thymic carcinoma are scarce, with only a few published reports describing the use of a drug-holiday approach for toxicity management (60,61). For now, such approaches should be considered an exploratory option for selected patients when standard dose reductions and brief interruptions fail to provide adequate toxicity control. ICI–related neuromuscular and cardiac toxicities ICIs show antitumor activity in thymoma and thymic carcinoma. However, unlike in many other solid tumors, their use in TETs is characterized by a distinctive pattern of irAEs, with neuromuscular and cardiac toxicities emerging as major concerns, especially in thymoma. Across recent prospective ICI-based trials in TETs (30,31,33,45,53,54,56), the incidence of grade ≥3 irAEs ranged from approximately 30% to 70% in thymoma and 5% to 15% in thymic carcinoma. Severe neuromuscular or cardiac irAEs were observed in approximately 30–40% of patients with thymoma but in up to 8% of those with thymic carcinoma. Despite this high burden of severe irAEs, treatment-related deaths were uncommon in those trials. However, these estimates (including mortality) are based on small patient cohorts, especially for thymoma, and should be interpreted with caution. Nevertheless, multiple fatal cases of overlapping myocarditis, myositis, and myasthenia gravis have been reported in case series, particularly in thymoma, underscoring the potential for catastrophic toxicity when these irAEs occur. A large review of 244 reported cases of ICI-related myocarditis found that concomitant myositis or myasthenia gravis was present in nearly half of patients (62). Additionally, a systematic review documented 60 cases of ICI-related myocarditis with overlapping myositis or myasthenia gravis across multiple tumor types, including TETs; in that cohort, the overlap manifestation typically developed after a median of one ICI dose and carried an in-hospital mortality rate of around 60% (63). Although thymic carcinoma shows a lower incidence of severe irAEs than thymoma, this difference is consistent with the well-recognized tendency of thymoma to be complicated by autoimmune diseases such as myasthenia gravis (64). Nevertheless, even in thymic carcinoma, the risk of ICI-related myotoxicities appears substantially higher than in other solid tumors; in a French nationwide study, the 6-month cumulative incidence of ICI-related myotoxicities reached 7.1% in thymic carcinoma, compared with an incidence of 0.9% across all tumor types (65). Taken together, these data illustrate that while the relative risk differs between thymoma and thymic carcinoma, both entities warrant particular caution when treated with ICIs. As indicated in the current NCCN guidelines (2), anti-PD-1/PD-L1 therapy is not recommended for thymoma. For thymic carcinoma, pembrolizumab is listed as a preferred regimen, with the caveat that PD-1/PD-L1 blockade in this entity may be associated with a higher incidence of irAEs. Given these substantial and early-onset risks, careful pretreatment evaluation is essential. Baseline assessment should include a careful history of overt or subclinical autoimmune disease and a focused neuromuscular examination. Moreover, baseline neuromuscular autoantibody screening (e.g., AChR antibodies) should be considered, particularly in patients with thymoma or when suggestive clinical findings are present. Cardiac evaluation should include a 12-lead electrocardiogram (ECG), and measurements of troponin and creatine kinase; baseline echocardiography is reasonable for patients with pre-existing cardiovascular disease or other high-risk features. Given that ICI-related myocarditis and myotoxicities typically manifest early, often within the first one or two cycles, close clinical reviews during the initial 4–6 weeks are crucial. For higher-risk patients, serial monitoring of cardiac biomarkers and ECG is essential, together with clear education for patients and caregivers to ensure prompt reporting of new cardiopulmonary or neuromuscular symptoms such as dyspnea, chest pain, ptosis, or limb weakness. At the first sign of myocarditis or neuromuscular-cardiac overlap, ICIs should be withheld, and the patient should be urgently hospitalized for a comprehensive multidisciplinary evaluation. Current cardio-oncology position statements (66) recommend the prompt initiation of high-dose corticosteroids (intravenous methylprednisolone, 500 to 1,000 mg per day) for severe ICI-related myocarditis, in parallel with ongoing diagnostic evaluation and clinical stabilization. They also advise early escalation to additional immunosuppression for steroid-refractory disease, or an upfront combination including abatacept, anti-thymocyte globulin, or intravenous immunoglobulin and plasma exchange for severe myocarditis or triple M syndrome (myocarditis-myositis-myasthenia). Permanent discontinuation of ICIs is recommended for any case of myocarditis grade 2 or greater. In clinical practice, ICIs should generally be avoided for thymoma outside clinical trials, except in highly selected patients after careful risk–benefit discussion and informed consent. In contrast, they are an accepted treatment option in previously treated thymic carcinoma, and emerging phase II evidence (e.g. the MARBLE trial) also supports first-line combination strategies; however, careful patient selection and vigilant safety monitoring remain critical. Future perspectives for advanced TETs In recent years, many clinical trials have been conducted to explore new systemic treatment approaches for advanced TETs. As summarized in Table 6, most studies are evaluating cytotoxic chemotherapy regimens, multitarget kinase inhibition, antibody-drug conjugates (such as agents targeting TROP-2), and combinations with ICIs in small phase II settings. These studies are expected to inform future treatment options, although their ultimate impact on standard care will depend on forthcoming results. Table 6Ongoing or planned clinical trials of systemic therapies for advanced TETsRegimenCytotoxic agentsTargeted agentsICIsOthersHistologyStudy designLineStudy (ref.)Carboplatin + nab-paclitaxel+TC, TPhase II1stjRCTs071240019 (67)PT-112+TC, TPhase II2nd+NCT05104736 (68)Rivoceranib+TC, TPhase IIAnyNCT06200233 (69)KC1036+TC, TPhase II2nd+NCT05683886Avelumab+TC, TPhase II2nd+NCT03076554Pembrolizumab + platinum + paclitaxel++TC, TPhase II1stNCT04554524Pembrolizumab + sunitinib++TCPhase II2nd+NCT03463460Tislelizumab + anlotinib++TC, TPhase II2nd+NCT06838910Lenvatinib + pembrolizumab + CP+++TC, TPhase II1stNCT05832827 (70)Sacituzumab govitecan-hziy++ADCTC, TPhase II2nd+NCT06248515Ivonescimab++Bispecific antibodyTCPhase II2nd+NCT06980077Bintrafusp alfa+Bifunctional fusion proteinTC, TPhase II2nd+NCT04417660ADC, antibody-drug conjugate; CP, carboplatin + paclitaxel; ICI, immune checkpoint inhibitor; ref., reference; T, thymoma; TC, thymic carcinoma; TET, thymic epithelial tumor. Ongoing or planned clinical trials of systemic therapies for advanced TETs ADC, antibody-drug conjugate; CP, carboplatin + paclitaxel; ICI, immune checkpoint inhibitor; ref., reference; T, thymoma; TC, thymic carcinoma; TET, thymic epithelial tumor. Alongside these therapeutic advancements, biomarker research has become increasingly important for refining patient selection and risk stratification for immunotherapy in TETs (Table S1). Although PD-L1 has been investigated as a potential biomarker in TETs, reported positivity shows substantial inter-study variation. This variability is partly due to differences in antibodies/assays, scoring systems, and intratumoral heterogeneity (71). Notably, given that thymic epithelial cells can naturally express PD-L1 as part of central tolerance programs, PD-L1 expression in TETs should be interpreted with caution. Consequently, PD-L1 alone may be insufficient as a predictive biomarker for ICI benefit (72,73). Furthermore, thymoma and thymic carcinoma are immunologically distinct; thymoma can maintain a thymus-like, immature thymocyte-rich environment, whereas thymic carcinoma more frequently exhibits a myeloid-skewed, immunosuppressive microenvironment (74,75). Recent genomic profiling of thymic carcinoma has highlighted recurrent alterations in key tumor suppressor genes and oncogenes (TP53, CDKN2A, CYLD, KIT) and several epigenetic regulators (TET2, SETD2, BAP1, and ASXL1). This analysis also confirmed that the tumor mutational burden is generally low, and microsatellite instability is rare. In particular, CYLD mutations have been proposed as a candidate biomarker for predicting response to ICIs (76), but the available data remain exploratory and require validation. Regarding toxicity, data from small thymoma cohorts and the broader literature on neuroimmune-related adverse events indicate that pre-existing autoantibodies, human leukocyte antigen (HLA) class I alleles, and baseline immune profiles may influence the risk of severe neuromuscular and cardiac toxicities (77,78). However, these potential biomarkers remain exploratory and largely unvalidated in thymic carcinoma. Looking ahead, composite biomarkers that integrate these genomic alterations with tumor- and host-immune features will be needed to better capture the likelihood of benefit and the risk of life-threatening immune-related toxicity in advanced TETs. Because TETs are rare, generating high-level evidence from large randomized phase III trials is challenging. Dedicated national and international infrastructure has therefore become crucial for building evidence. In France, the RYTHMIC network operates as a nationwide prospective program that includes multidisciplinary tumor board discussions and data collection (79,80). The International Thymic Malignancy Interest Group (ITMIG) and its collaborative databases have also demonstrated the feasibility of large-scale, coordinated, global data pooling (81). Integrating future systemic therapy trials and translational research into these platforms is crucial for effective validation of new treatment strategies and biomarkers for TETs. Artificial intelligence (AI) is emerging as a key tool to address the diagnostic and prognostic complexity of TETs, mainly through imaging- and pathology-based biomarkers. A recent systematic review identified 65 AI-based studies on TETs (82). Across these and other studies, computed tomography (CT)-based radiomics and deep learning models can assist in tumor characterization and WHO risk classification (83,84), and digital pathology approaches using deep learning and registry-based prediction models can support histologic subtyping and survival prediction (85-87). However, most existing models have been developed retrospectively, in single-center cohorts, and outside the framework of prospective therapeutic trials. Future research should focus on prospective, multi-institutional validation of AI-derived imaging and pathology signatures and their integration into trials of systemic therapy for advanced TETs. The international INTHYM study (NCT06301945) is currently ongoing to develop and validate AI-based tools for histologic assessment and recurrence risk prediction in TETs. Any AI-based biomarker entering clinical decision-making should meet emerging methodological standards, such as the European Society for Medical Oncology (ESMO) Basic Requirements for AI-based Biomarkers in Oncology (EBAI) (88). Conclusions Systemic treatment for advanced TETs continues to include cytotoxic chemotherapy as a major component. Accumulating evidence suggests clinically meaningful activity of targeted agents and ICIs in selected patients and supports the ongoing development of combination strategies. The emergence of novel agents, such as antibody-drug conjugates and bispecific T-cell engagers, as well as the integration of immunotherapy and targeted therapy, is expected to expand treatment options. Prospective collaborative trials with rigorous safety monitoring and biomarker-based patient selection are needed to define optimal treatment strategies and maximize clinical outcomes. Supplementary The article’s supplementary files as Available at https://med.amegroups.com/article/view/10.21037/med-2025-1-77/prf