Immunotherapy and Targeted Therapies Efficacy in Thymic Epithelial Tumors: A Systematic Review

Background: Thymic epithelial tumors (TET) are rare neoplasms of the anterior mediastinum. Surgery is the mainstay treatment for resectable TET, whereas systemic treatments are reserved for unresectable and metastatic tumors. The development of new treatments, such as immune checkpoint inhibitors (ICI) and targeted therapies, with promising results in other types of solid tumors, has led to the investigation of their potential efficacy in TET. The study of tumor microenvironments (TME) is another field of investigation that has gained the interest of researchers. Taking into account the complex structure of the thymus and its function in the development of immunity, researchers have focused on TME elements that could predict ICI efficacy. Materials and Methods: The primary objective of this systematic review was to investigate the efficacy of ICI in TET. Secondary objectives included the toxicity of ICI, the efficacy of targeted therapies in TET, and the evaluation of the elements of TME that may be predictive factors of ICI efficacy. A literature search was conducted in February 2023 using the Ovid Medline and SciVerse Scopus databases. Results: 2944 abstracts were retrieved, of which 31 were retained for the systematic review. Five phase II and one retrospective study assessed ICI efficacy. The overall response rate (ORR) varied from 0% to 34%. Median progression-free survival (PFS) ranged from 3.8 to 8.6 months, being lower in thymic carcinoma (TC) (3.8–4.2 months). Median overall survival (OS) ranged from 14.1 to 35.4 months. Treatment-related adverse events occurred in 6.6% to 27.3% of patients. Sixteen studies assessed targeted therapies. The most active molecule was lenvatinib, with 38% ORR in patients with TC while no activity was detected for imatinib, erlotinib plus bevacizumab, and saracatinib. Ten studies assessed TME elements that could predict ICI efficacy. Four studies focused on the tumor-infiltrating immune cells suggesting improved outcomes in patients with TC and high tumor-infiltrating lymphocyte densities. Another study showed that CD8+, CD20+, and CD204+ tumor-infiltrating immune cells in cancer stroma might be prognostic biomarkers in TC. Another study identified the immune-related long non-coding RNAs as a predictor of response to ICI. Tumor mutational burden was identified as a predictive factor of ICI efficacy in one study. Conclusions: Despite study heterogeneity, this review shows that ICI could be a therapeutic option for selected patients with TET that are not amenable to curative radical treatment after first-line chemotherapy.


Introduction
Thymic epithelial tumors (TET), which account for 15% of all anterior mediastinal tumors, are uncommon neoplasms of the prevascular mediastinum [1].They are derived from the epithelial cells of the thymus and are categorized in relation to the fraction of the non-tumoral lymphocytic part, and to their resemblance to normal thymic architecture.This heterogeneous group of neoplastic lesions includes thymomas and thymic carcinomas (TC).The 2015 revised World Health Organization (WHO) classification system classified TET as A, AB, B1, B2, and B3 thymoma, and TC [2].Thymomas may present an indolent course and, for that reason, were formerly considered benign neoplasms.However, they are nowadays classified as malignant lesions.The overall prognosis is good for thymomas that are amenable to complete surgical resection [3].On the contrary, TC are characterized by a more clinically aggressive behavior [3,4].The majority of patients are eligible for surgical treatment, which may be combined with adjuvant radiation.Chemotherapy is prescribed to patients with unresectable TET, at advanced stages (stages III-IV according to either the Masaoka-Koga or the ITMIG classification), or for recurring diseases [4,5].Platinum-based chemotherapy is the most often used regimen but there is no formal consensus on the best regimen due to the absence of randomized trials in these rare entities [3][4][5].
Similarly, there is an increased interest in the detection of molecular alterations that could be modified by targeted therapies [14][15][16][17][18][19][20][21][22][23][24][25][26][27][28][29].New drugs are selectively targeting the pathways that play an important role in oncogenesis, tumor growth, and proliferation.The rarity and the histological heterogeneity of TET constitute a major obstacle in the conduct of large-scale trials and, until now, the majority of evidence has derived from case reports and small case series.
Another research field that has gained recent interest is that regarding the tumor microenvironment (TME) [30].The TME consists mainly of endothelial cells of the vascular epithelium, cells of innate and adaptive immunity, fibroblasts, pericytes, signal-carrying molecules, and the extracellular matrix.A close dynamic relationship exists between the tumor and its microenvironment [31][32][33].This complex interaction is a key element in oncogenesis, growth, and tumor spread.Taking into account the complex structure of the thymus and its function in the development of immunity, researchers are focusing on the TME elements that could predict ICI efficacy [32,33].The expression of PD-L1 on tumor cells contributes to the prediction of clinical efficacy of ICI in some tumor types, such as non-small cell lung cancer, which take advantage of the crucial roles played by the PD-L1/PD-1 and CTLA4/CD80/CD86 axes in the evasion of immune surveillance.As a result, the distinct composition of TME within various TET histological categories, along with the variations seen among them in the PD-L1 expression patterns, constitutes an essential component of their biological background and directly regulate both the response to checkpoint-inhibitory receptor blockade and the predisposition to autoimmune disease [31][32][33].
The objective of this systematic review was to assess the most recent data on the effectiveness of immunotherapy treatments and the advantages of targeted therapies against potentially treatable mutations in TET.Additionally, components of the TME that might serve as indicators of ICI effectiveness were evaluated.

Material and Methods
The literature search, which was designed by a scientific librarian (VD) with competence in medical literature research, was carried out in February 2023 utilizing the Ovid Medline and SciVerse Scopus databases.The search parameters were converted into MeSH terms and free-text keywords, which were then used to search for specific information in titles, abstracts, keywords, and substance names (where applicable) in Medline and titles, abstracts, and keywords in Scopus (Appendix A).The resulting citations were exported from Medline and Scopus into a reference manager software (EndNote version X9) to remove any duplicates, and then in a dedicated systematic literature reviews system (https://rayyan.ai,accessed on 2 February 2023) for the selection process.The researchers worked simultaneously by composing two pairs (ACA and CJ, TB and MB) that ran the initial article selection independently.In the first step, articles were selected if deemed eligible based on the title and abstract content.The final selection was performed after reading the full-text article.The selections of both groups of reviewers were compared and the mutually selected papers represented the total eligible papers to be analyzed.Any discrepancies between the two groups of researchers were resolved after a consensual discussion.The selected articles' references were examined to detect any missing potentially eligible publications.
The inclusion criteria were the following: Only articles in English, French, or Dutch were considered.There was no selection based on the year of publication.
(1) Phase II/III clinical trials and retrospective series (>14 patients according to Simon's design) [34,35] assessing ICI in TET and reporting at least one of the following clinical outcomes: a. progression-free survival (PFS), defined as the time from randomization to disease progression or death from any cause; b.
overall survival (OS), defined as the time from randomization until death from any cause; c.
objective response rate (ORR), defined as the proportion of patients who achieved an objective response (partial or complete according to the Response Evaluation Criteria in Solid Tumors (RECIST)); d.
all grade or grade ≥ 3 treatment-related adverse events.(2) Phase I/II/III clinical trials and retrospective series (>14 patients according to Simon's design) assessing targeted therapies against an oncogenic driver mutation or translocation (EGFR, cKIT, KRAS, ALK, BRAF, PDGFR, HER2, MET etc.).(3) Experimental cohort studies investigating any of the following: − TME of TET, % of PD-L1 expression in TET or tumor mutational burden (TMB) AND prediction of ICI efficacy.
Phase I trials concerning different types of tumors, even including TET, were not considered.
The following data were extracted: study characteristics (design, patient selection), patients' characteristics (gender, age, previous treatments), stage and histology, treatments and clinical outcomes (number of arms, hazard ratio (HR), and 95% confidence interval (CI) for PFS and OS, overall response rate (ORR), number of patients with grade 3 or greater adverse events (AEs).
The main judgment criterion was ORR.Secondary judgment criteria were PFS, OS (median and at specific time-point: 1-2-5 years), and grade 3-5 AEs.

Data Synthesis
Given the high heterogeneity in the selected studies in terms of inclusion criteria, treatments, and data presentation, a quantitative analysis was not performed.

Results
A total of 2944 abstracts were retrieved through the search equation, of which 31 were eligible for the systematic review.The PRISMA flowchart is depicted in Figure 1.
Six trials assessed ICI efficacy in TET [9,13,[36][37][38][39][40][41] (Tables 1 and 2).Five were phase II trials whereas the last was a retrospective cohort with 77 patients enrolled.All were recently published, from 2018 to 2023.Four studies were multicentric and two were elaborated in a single center.The administered drug was pembrolizumab in two studies and nivolumab in another two studies.In one study the ICI avelumab was combined with the anti-angiogenic agent axitinib.In the last study, different PD-1 inhibitors (anti-PD-1: nivolumab, pembrolizumab, sintilimab, camrelizumab, tislelizumab, and toripalimab) were used.The median number of patients was 37 (range .ORR was the primary endpoint of four phase II trials and the PFS rate at six months of the fifth one.In half of the studies, only patients with TC were included whereas patients with thymomas and TC were assessed in the remaining studies.The Masaoka-Koga classification was used throughout the trials.All patients presented with stage III (which were not candidates for curative surgical resection) or IV (IVa and/or IVb).The median follow-up duration was 14.9 months (range 13.3-22.4months).The ORR varied from 0% to 34%.In trials exclusively enrolling patients with TC, the ORR was 0% to 22.5%.The mPFS ranged from 3.8 to 8.6 months overall, being 3.8 to 4.2 months in TC.The mOS ranged from 14.1 to 35.4 months.Treatment-related AE occurred in 6.6% to 27.3% of patients (Table 2).
Sixteen studies testing targeted therapies were deemed eligible for further analysis [23,25,[42][43][44][45][46][47][48][49][50][51][52][53][54].There were thirteen phase II trials, two retrospective studies, and one prospective cohort (Tables 3 and 4).There were four single-center studies, five studies were conducted in two centers and the remaining seven studies were multicenter trials.Different molecules with various actions were evaluated.Lenvatinib, a multi-targeted inhibitor of VEGFR, FGFR, RET, c-Kit, and other kinases.Publication years ranged from 2008 to 2023, and the number of enrolled patients varied from 14 to 72.The median follow-up duration ranged from 15.5 to 46 months.Most of the studies (13 out of 16) enrolled patients with both thymomas and TC.In all studies, patients were treated with at least one prior chemotherapy scheme treatment.In seven studies, the ORR was the primary endpoint and ranged from 0% to 38%.The highest ORR (38%) was observed in patients with TC treated with lenvatinib, and mOS was not reached.In three studies with imatinib, a combination of erlotinib and bevacizumab, and saracatinib, the ORR was 0%.Grade 3-4 treatment-related AEs varied according to the molecule and are reported in Table 4.
Ten studies assessed TME elements that could predict the efficacy of ICI [55][56][57][58][59][60][61][62][63][64].One study was published in 2011 and the others from 2019 to 2022.Three studies were based on resected TC specimens (10, 32, and 39 patients).Four studies included mixed histologies (thymomas and TC) with 21, 22, 31, and 33 specimens.In three studies, TETs were included among different cancer types.The results of these different studies are detailed in Table 5.Four studies focused on the tumor-infiltrating immune cells suggesting improved outcomes in patients with TC and high tumor-infiltrating lymphocyte densities [55,57,59,60].Another one showed that CD8+, CD20+, and CD204+ tumor-infiltrating immune cells in cancer stroma might be prognostic biomarkers in TC [58].Another study identified the immunerelated long non-coding RNAs as a predictor of response to immunotherapy [56].A Chinese study identified TMB as a predictive factor of ICI efficacy [63].Researchers from South Korea assessed the first-week proliferative response of PD-1+ CD8+ T cells as a predictive marker [61].The immunological pathways predisposing to irAE are the subject of another study [62].Six trials assessed ICI efficacy in TET [9,13,[36][37][38][39][40][41] (Tables 1 and 2).Five were phase II trials whereas the last was a retrospective cohort with 77 patients enrolled.All were recently published, from 2018 to 2023.Four studies were multicentric and two were elaborated in a single center.The administered drug was pembrolizumab in two studies and nivolumab in another two studies.In one study the ICI avelumab was combined with the anti-angiogenic agent axitinib.In the last study, different PD-1 inhibitors (anti-PD-1: nivolumab, pembrolizumab, sintilimab, camrelizumab, tislelizumab, and toripalimab) were used.The median number of patients was 37 (range 15-77).ORR was the primary endpoint of four phase II trials and the PFS rate at six months of the fifth one.In half of the studies, only patients with TC were included whereas patients with thymomas and TC were assessed in the remaining studies.The Masaoka-Koga classification was used throughout the trials.All patients presented with stage III (which were not candidates for curative surgical resection) or IV (IVa and/or IVb).The median follow-up duration was 14.9 months (range 13.3-22.4months).The ORR varied from 0% to 34%.In trials exclusively enrolling patients with TC, the ORR was 0% to 22.5%.The mPFS ranged from 3.8 to 8.6 months overall, being 3.8 to 4.2 months in TC.The mOS ranged from 14.1 to 35.4 months.Treatment-related AE occurred in 6.6% to 27.3% of patients (Table 2).

Discussion
TETs are rare and histologically heterogeneous tumors.However, they represent the majority of neoplasms located in the anterior (prevascular) mediastinum.Surgical resection is the treatment of choice for early-stage and resectable tumors, whereas there is no consensus about the best systematic treatments for advanced diseases.This systematic review presents updated data on the clinical activity of ICI and targeted therapies in advanced and metastatic thymoma and TC.It provides a comprehensive review of the existing evidence, which has, until now, been represented mainly by anecdotal case reports and small case series.

Immune Checkpoint Inhibitors in TET
Immunotherapy has drawn the attention of researchers who are looking into its effectiveness in TET because it has demonstrated encouraging effects in other solid tumors.The crucial function of the thymus gland in the formation of adaptive immune responses makes the TET example fascinating [65].ICI enhances the immune response against tumors but may also trigger immune-related adverse events (irAE).Accordingly, the activity and toxicity of ICI in clinical trials are emphasized in this systematic review.Preliminary encouraging clinical results are reported with ORR up to 34%, mPFS ranging from 3.8 to 8.6 months, and mOS between 14.1 and 35.4 months.A list of ongoing trials derived from clinicaltrials.gov is shown in Table 6.
However, substantial toxicity is observed in 6.6% to 27.3% of patients, which is a significant barrier to their routine and widespread use.IrAEs generally carry a tolerable level of morbidity but occasionally result in fatal outcomes (varying from 0.36% to 1.23%).Lethal toxicities are less frequent with anti-PD-1 and anti-PD-L1 antibodies, in comparison with anti-CTLA-4 antibodies and certainly for combined administrations (anti-PD-1/PD-L1 plus anti-CTLA-4) [36,66,67].Currently, ICI (pembrolizumab, nivolumab, and avelumab) must only be used in clinical studies, as the most appropriate group of patients, those with reduced risk of irAEs and with the best therapeutic benefit, must yet be defined.
Immunotherapy agents, unlike those used in conventional cytotoxic therapy, work therapeutically by inducing the anti-tumor immune response, which is based on the immunoregulative process that takes place between cancer cells and the TME.In numerous cancer types, attempts to link PD-L1 expression in tumor cells and ICI effectiveness have shown inconsistent results [58].PD-L1 expression in thymic epithelial malignancies has been documented in earlier research, but its application as a diagnostic biomarker in TET is not well understood.Therefore, more precise biomarkers and more pertinent predictive features for the identification of individuals who will potentially benefit from ICIs are needed to guide patients' selection [55].
The thymus is an organ in charge of the evolution of adaptive immunity.Thymus cell lymphocytes or T cells, which are crucial components of adaptive immune function, mature in the thymus.More specifically, the complex thymic structure provides a special microscopic environment that directs thymocyte maturation and instructs T cells to develop self-tolerance [65].More lymphocytes may boost the likelihood of a successful application of ICI because they fight cancer by increasing cytotoxic lymphocytes [59].ICI effectiveness in TC patients is supported by improved outcomes in patients with high tumor-infiltrating lymphocyte densities [60].Effector cells targeting cancer cells, including CD8+ cells, are a predictive marker for ICI activity [68].The balance between effector and suppressor cells may be crucial for the TME function and might serve as a prognostic and predictive biomarker for ICI [69].Previous studies have suggested that effector cells, such as CD8+ lymphocytes, are favorable prognostic indicators among patients with TC, but these data are inconsistent [57,70].Sato et al. have demonstrated that CD8+, CD20+, and CD204+ tumor-infiltrating immune cells in cancer stroma might be prognostic biomarkers in TC.More specifically, high mean numbers of stromal CD8+, CD20+, and FOXP3+ cells have been shown to be significantly associated with favorable prognosis, whereas high CD204+ cell density tended to be correlated with poor prognosis [58].Shim et al. have confirmed these findings, with a link between better TC prognosis and higher density of stromal CD20+ cells (B lymphocytes) [57].These results suggest that thymic malignancy differs from other cancer types in the influence of CD20+ cells and that the density of CD20+ tumor-infiltrating immune cells in stromal lesions has to be examined.This opens the door to the investigation of immunotherapy approaches targeting B cell immunity in TC [57].Other arguments suggest the important role of TME in predicting ICI activity or toxicity in TET.Su et al. have developed an immune-related long noncoding RNAs classifier to pinpoint the response in patients with TET.As the authors advocate, long noncoding RNAs can control the immune response by controlling homeostasis, TME, anti-inflammatory agents, and immune cell activity [56].Six prognosis-related immune-related long noncoding RNAs (AC004466.3,AC138207.2,AC148477.2,AL450270.1,HOXB-AS1, and SNHG8) were selected to build an immune-related long noncoding RNAs classifier.According to these authors, their model can be used to forecast outcomes, the degree of immune infiltration, and the effectiveness of immunotherapy in patients diagnosed with TETs.It may also help with individualized immunotherapy counseling.
Kim et al. assessed the first-week proliferative response of PD-1+ CD8+ T cells as a predictive marker of tumor responses to anti-PD-1 therapy and clinical outcomes in patients with TETs.The proliferative response after anti-PD-1 therapy was evaluated by the fold-change in the percentage of Ki-67+ cells among PD-1+ CD8+ T cells on day 7 (Ki-67 D7/D0).In the cohort of patients with TETs, Ki-67 D7/D0 was found to be significantly higher in patients with durable clinical benefits than in those with no durable benefits [61].However, Ki-67 D7/D0 significantly predicted OS in patients with non-small cell lung cancer, but not in patients with TETs.
Chen et al. investigated the association between protein kinase, DNA-activated, catalytic subunit (PRKDC) mutations and TMB, TME, and response to ICI on solid tumor samples collected from 3877 patients that underwent a panel-based next-generation sequencing assay [63].PRKDC is an important gene for DNA double-strand break repair and central T-cell tolerance.PRKDC mutation is one of the significant factors linked to increased TMB, inflamed TME, and greater responsiveness to ICI.It frequently appears to co-exist with defects in other DNA damage repair pathways.However, their specificity in TET needs to be validated in larger-scale trials.
Yip1 interacting factor homolog B (YIF1B) is a membrane protein that belongs to the FinGER protein family.It is involved in the endoplasmic reticulum (ER)-to-Golgi trafficking [71].Recent research has demonstrated its role in serotonin-induced cancerogenesis.Liu et al. found a positive relationship between YIF1B expression and immune cell infiltration in several cancer types, and YIF1B expression was also found to be positively correlated with TMB, microsatellite instability, and methylation in some cancer types, linking its expression to a possible evaluation of therapy response [64].
In another study, peripheral blood T-cell characteristics are linked to the emergence of irAEs following anti-PD-1 medication and four different patient subgroups are defined: Th17-related, TNF-related, CD8-related Treg-compensated, and CD8-related Treguncompensated.Patients with severe irAEs presented a significantly lower fold increase in the frequency of effector regulatory T (eTreg) cells after anti-PD-1 treatment, a higher proportion of T helper-17 (Th17) and T helper-1 cells in the beginning, and an increased fraction of Ki-67+ cells among PD-1+ CD8+ T cells post treatment.Various irAE subtypes may have unique underlying immunological processes [62].Early assessment of immune responses may also have clinical implications for irAE prediction.

Targeted Therapies in TET
Compared with thymomas, TC exhibits more somatic mutations in cancer-related genes [72].Thus, it is reasonably expected that thymoma and TC may have distinct responses to targeted therapies.Different potential targets have been identified in thymoma and/or TC that are reported hereafter.
Except for a few isolated case reports in Asian individuals, somatic activating EGFR mutations are relatively uncommon in thymic malignancies [73,74].EGFR protein overexpression is present in approximately 70% of thymomas and 50% of TC without any relationship with the histologic subtype [75,76].About 20% of thymic malignancies exhibit EGFR gene amplification by fluorescence in situ hybridization (FISH), most frequently in type B3 thymoma and TC, related to more advanced stage and capsule invasion.
KIT immunohistochemical positivity can be seen in up to 73-86% of TC but only in 2% of thymomas [76,77].As KIT is a target in other tumor forms, most notably in gastrointestinal stromal tumors, this variation in tumor biology results in a definite difference in therapeutic approaches between TCs and thymomas.Unfortunately, the rate of KIT mutations is still only 7 to 9%, despite the high frequency of KIT expression in TC.Four mutations have been described to date: the V560 deletion and L576P substitution found in exon 11, the D820E mutation in exon 17, and the H697Y mutation found in exon 14 [20,22,78].
Angiogenesis significantly influences TET carcinogenesis.Both thymomas and TC overexpress vascular endothelial growth factor (VEGF)-A and VEGFR-1 and -2, although there is little information on the effectiveness of angiogenesis inhibitors in thymic malignancies [79,80].Low response rates have been observed with bevacizumab [21].The activity of multikinase inhibitors, particularly sorafenib, and sunitinib, has been emphasized in case reports involving TC [21].While multikinase inhibitors may have some impact on TC, angiogenesis inhibitors by themselves do not seem to have an effect on either thymomas or TC.
Histone deacetylase (HDAC) inhibitors, in particular the pan-HDAC inhibitor belinostat, were also assessed in TET.A patient with thymoma who participated in a phase I study with belinostat experienced a mild response that lasted for 17 months [83].A phase II trial with intravenous infusion of belinostat showed only two partial responses in thymomas (ORR 8%, 95% CI 2.3-25.9%)and no responses in TC [25].
The present systematic review has pointed out that the most active agent is lenvatinib, as demonstrated in the REMORA phase II trial.Lenvatinib is a multi-targeted inhibitor of VEGFR, FGFR, RET, c-kit, and other kinases.Further, we can suggest sunitinib as an acceptable second-line therapy for TC [44].Imatinib demonstrated limited activity in chemotherapy-pretreated patients with TC harboring KIT mutations [27,[84][85][86].Everolimus is a potential treatment option for pre-treated patients with TETs when considering durable disease control in a significant proportion of patients with thymomas or TC [49].Further investigations are underway, including antiangiogenic combinations, for example, ramucirumab with carboplatin and paclitaxel in a first-line setting [87].

Table 1 .
List of the studies assessing ICI in TET (study design and patient enrollment).ICI: immune checkpoint inhibitors, M: male, F: female.

Table 2 .
List of the studies assessing ICI in TET (objectives and outcomes).

Table 3 .
List of the studies assessing targeted therapies in TET (study design and patient enrollment).

Table 4 .
List of the studies assessing targeted therapies in TET (objectives and outcomes).

Table 5 .
Studies investigating the elements of the TME that could be predictive factors of ICI efficacy.

Table 6 .
List of ongoing trials assessing immune checkpoint inhibitors in thymic epithelial tumors.