Thymic epithelial tumors (TET, thymomas, and thymic carcinomas) are the most frequent neoplasias originating from the thymus. The fourth edition of World Health Organization (WHO) classification of thymic tumors recognizes 10 histologic categories of thymomas and 10 further categories of thymic carcinomas.1 Previous studies established that radical surgery is the most effective therapeutic approach in thymic epithelial tumors.2 Multivariate analysis proved that invasiveness at the tumor resection margin is one of the most important factors indicating adverse prognosis.3,4 In case of advanced or recurrent disease, platinum-based conventional chemotherapy and/or perioperative radiotherapy can be administered. Recently, different tyrosine kinase or growth factor receptor inhibitors were also investigated as an alternative therapeutic option in progressive cases.5,6 However, because of the low frequency of indicative biomarkers, for example, activating mutations, these studies showed inconclusive results so far.7,8
The manipulation of immune resistance checkpoints, first of all the blockage of the interaction between programmed cell death receptor (PD-1) and its ligand (PD-L1) seems to be a promising therapeutic target in malignant tumors.9–11 For the administration of a certain PD-1/PD-L1 inhibitors approved by the Food and Drug Administration (FDA), PD-L1 expression of the tumor cells or tumor-infiltrating immune cells should be tested by a complementary FDA-approved diagnostic test based on immunohistochemistry and a special evaluation scheme. Atezolizumab, for example has recently received FDA-approval for use in a certain group of patients with metastatic nonsmall cell lung cancer (NSCLC) along with the Ventana PD-L1 (SP142) assay as a complementary diagnostic test.12 However, assays established for a given histologic condition require specific clones, immunohistochemistry protocols, and scoring systems that make the use highly difficult in a different rare neoplasia to support therapeutic alternative, such as in case of thymic tumors.13 Although PD-L1 expression in TETs has been published before, we miss information about the utility and expression level of Ventana PD-L1 (SP142) related approach in these tumors.14,15 This particular procedure seems to have a high clinical impact for treatment planning specifically with PD-L1 inhibitor atezolizumab in isolated aggressive cases or trials focusing on thymic neoplasias.
In this study, we provide information about PD-L1 expression in tumor cells and tumor infiltrating immune cells of 36 thymic epithelial tumors (29 thymomas and 7 thymic carcinomas) using primary monoclonal antibody Ventana PD-L1 clone SP142 and the recommended evaluation scheme provided by the supplier. In addition, the PD-1 status and the mutational status of common oncogenes including EGFR, KRAS, and BRAF were also determined and compared.
MATERIALS AND METHODS
Patients and Samples
A total of 36 newly diagnosed TET cases presented in the Clinical Centre of the University of Debrecen between 2005 and 2015 were retrospectively evaluated. The mean age was 53.38 years ranging between 16 and 79 years, female to male ratio was 1:1; 7 of 36 cases were diagnosed with thymic carcinoma (4 squamocellular carcinoma; 1 NUT midline carcinoma; 1 adenocarcinoma; 1 lymphoepithelioma like carcinoma) and 29 cases were diagnosed with thymoma (3 type A, 7 type AB, 6 type B1, 9 type B2, 3 type B3, 1 micronodular thymoma). Tumor resection was performed in all cases. In 8 cases, perioperative radio-based and/or cisplatin-based chemotherapy was additionally administered. In 1 thymic carcinoma with disease progression cetuximab therapy was used. The median follow-up period was 26 months (ranging between 0 and 125; mean, 36.79 mo). The study design was in agreement with the local ethical and patient safety standards.
Immunohistochemical (IHC) and Molecular Studies
Tissue sections of 4 μm thickness from representative formalin-fixed paraffin-embedded tissues were used for serial H&E and IHC staining. Percentage of lymphoid component of the tumor was evaluated in all cases based on H&E-stained slides.
Deparaffination, antigen retrieval, staining for PD-L1 (clone: SP142, 07309457001, Spring Bioscience Corp., Roche Diagnostic Hungary) and PD-1 (clone: NAT105, ACI3137C, Biocare, Biocare Medical Inc.) was performed using the Ventana Ultra automated IHC stainer according to the instructions of the manufacturer. For the validation of the original diagnosis pCK (clone: MNF116, M0821, Agilent Pathology Solution, Kromat Ltd, Hungary), TdT, CD117 (clone: EP10, CME296C, Biocare Medical Inc., Frank Diagnosztika Ltd., Hungary) IHC examination was performed.
Baseline PD-L1 expression was scored following immunohistochemistry in the tumor cells (as percentage of tumor cells showing PD-L1 membrane staining of any intensity, TC value) and tumor-infiltrating immune cells (as percentage of tumor area occupied by positive immune cells of any intensity, IC value) as described previously.16 At least 1% expression rate was considered as positive for both parameters. All cases were categorized based on approved cut-off values (TC≥50% for tumor cell and IC≥10% area for tumor infiltrating immune cells). All of the applied categories were originally introduced for NSCLC cases tested with the same Ventana PD-L1 clone SP142, where PD-L1 expression ≥50% TC or ≥10% IC was found to be associated with atezolizumab-related enhanced overall survival.16 In lymphocyte rich histologic subtypes, such as type B1, B2, and AB thymoma, pan-cytokeratin staining was performed on a serial slide to highlight the epithelial component (Fig. 1).
PD1 expression was evaluated in tumor-infiltrating immune cells only as tumor cells were negative for PD1 staining in all case. Cases were categorized into 3 groups based on the percentage of tumor area occupied by PD1 positive immune cells: PD1 negative <1%; PD1 low-expression (≥1% and <5%), PD1 high-expression (≥5%). To demonstrate the exact location and distribution of PD1 positive immune cells within the samples PD1/pCK (pCK clone: MNF116, M0821, Agilent Pathology Solution, Kromat Ltd., Hungary) and PD1/CD23 double immunohistochemistry was performed in selected cases. VIP chromogen was used to visualize the second primary antibody in case of double staining.
All immunohistochemistry parameters were determined in a Leica DM2000 microscope and cross-checked by an independent second pathologist for interobserver variations.
EGFR exon 19 and exon 21, KRAS exon 2, BRAF exon 15 mutational status was analyzed in cases with thymic carcinomas. Molecular testing was performed retrospectively using formaldehyde-fixed and paraffin-embedded tissue blocks. Following DNA isolation and PCR-amplification conventional Sanger sequencing was done to identify common mutational hotspots.
Statistical evaluation and graphs were made using GraphPad Prism Version 5.03 software. Mean and SDs were calculated for each group and analyzed with Fisher exact test, χ2 test or Mann Whitney test. P-values <0.05 were considered statistically significant. Correlations between data sets were obtained using the Spearman test and linear regression analysis.
Evaluation of PD-L1 IHC Staining in Thymic Epithelial Tumors
Altogether 36 TET cases were tested for PD-L1 expression. Table 1 summarizes the most important biological features of our cases such as sex, age, WHO histologic subtype, Masaoka-Koga stage, TNM stage, and size. Information about the clinical stage was not available in case of one patient (foreign citizen, treatment details not available). PD-L1 expression of tumor cells were evaluated together with pan-cytokeratin staining that highlighted the epithelial component. Tumor cells showed continuous or partial membrane PD-L1 expression (Fig. 1). Although some thymic epithelial tumors contained a huge amount of lymphocytes, PD-L1 expression was primarily detected mostly in macrophages or dendritic cells among tumor infiltrating tumor cells, whereas the lymphocytic infiltrate remained negative (Fig. 2).
PD-L1 Expression of Tumor Cells and Tumor Infiltrating Immune Cells
PD-L1 TC could be evaluated in all cases, the inter-reader agreement was very good (weighted κ: 0.840) and the correlation between the observers was significant (Spearman r=0.966; P<0.0001). PD-L1 IC could be detected in 35 of 36 cases, in one case with thymic carcinoma nonspecific positivity was observed in granulocytes. The inter-reader agreement was good (weighted κ: 0.6788) and the correlation between the observers was also significant (Spearman r=0.9015; P<0.0001).
Tumor cells were completely negative for PD-L1 in 10 of the 36 cases, in 9 of 29 thymomas (31.0%) and in 1 of 7 thymic carcinomas (14.3%). Further, 11 cases were positive but did not reach the cut off value (TC <50%) consisted of 6 of 29 thymomas (20.7%) and 5 of 7 thymic carcinomas (71.4%). High rate PD-L1 expression (TC ≥50%) was seen in 15 cases, including 14 of 29 thymomas (48.3%) and 17 thymic carcinomas (14.3%) (Fig. 3). Altogether one thymic carcinoma was negative for PD-L1 TC (NUT-midline carcinoma) and one case showed high rate PD-L1 expression with TC≥50% (squamocellular carcinoma).
PD-L1 positive tumor infiltrating immune cells were completely lacking in 4 of 35 cases (11.4%), made of 4 of 29 thymomas (13.8%) and 0 of 7 thymic carcinomas (0.0%). In 17 cases (48.6%) PD-L1 positive tumor infiltrating immune cells could be detected in <10% of tumor area (IC<10%), this was observed in 13 of 29 thymomas (44.8%) and 4 of 7 thymic carcinomas (57.1%). Moreover, IC was ≥10% in total of 14/35 cases (40.0%) including 12/29 thymomas (41.4%) and 2/7 thymic carcinomas (28.6%) (Fig. 3). One thymic carcinoma with high IC value was a squamocellular carcinoma, whereas the other one was lymphoepithelioma-like carcinoma.
Altogether 20 of 29 thymomas (69%) and 6 of 7 thymic carcinomas (86%) were positive for PD-L1 (TC ≥1%), but only 9 of 29 thymomas (31%) and 1 of 7 thymic carcinomas (14%) have reached 50% positivity (TC≥50%). Considering TC and IC values together 20 of 29 thymomas (69%) and 3 of 7 thymic carcinomas (43%) showed ≥50% TC or ≥10% IC, which values indicate anti-PD-1/PD-L1therapeutic advantage in NSCLCs when the same conditions are applied (Fig. 3).
Correlation Between PD-L1 Expression and Histologic Subtypes of TETs
There was no significant difference between thymomas and thymic carcinomas in the percentage of PD-L1 positive tumor cells (mean TC value 40.17 vs. 25.00, respectively; P=0.5018) nor in the percentage of PD-L1 positive immune cell-occupied tumor area (mean IC value 8.45 vs. 11.50, respectively; P=0.8079).
There was no significant connection between the WHO histologic subtype and the percentage of PD-L1 positive tumor cells (TC value) or the percentage of PD-L1 positive immune cell-occupied tumor area (IC value) (P=0.0721 and P=0.5816 respectively, χ2 test). Also, there was no correlation between TC and IC values within the same tumor (P=0.3219; Spearman r=−0.1699).
Assuming that spindle and polygonal cells forming A-type and B-type thymomas have different cell origin, we artificially divided thymomas into 2 groups (type A, AB vs. B1-3). It is interesting to note that, in A and AB type (medullary) thymomas significantly lower amount of PD-L1 positive tumor cells could be detected as in type B1-3 thymomas of pure cortical epithelial origin (P=0.0488; mean 8.00 vs. 51.94, respectively). In the subgroup of medullary origin, 2 outliers (one A- and one AB type case) could be observed with high percentage of PD-L1 positive tumor cells.
Correlation Between PD-L1 Positive Immune Cell-occupied Tumor Area (IC Value) and the Lymphoid Component of Thymic Tumors
As thymic epithelial tumors contain a large amount of thymic lymphoid cells depending on the subtype the connection between the percentage of the lymphoid component and the IC value was analyzed in detail. Not surprisingly, there was significant connection between the histologic subtype and the number of lymphoid cells using ANOVA test (P=0.0002), type B1, B2, an AB thymoma contained the highest percentage of lymphoid cells. There was a weak but significant correlation between the percentage of lymphoid cells and the percentage of PD-L1 positive IC value (P=0.0297; Spearman r=0.3627). As already stated, there was no significant connection between the WHO histologic subtype and IC value. It should be emphasized that PD-L1 positive immune cells were mostly enlarged cells with broad cytoplasm representing the macrophage and dendritic cell lineage, whereas masses of compact lymphoid cells were negative. It can be assumed that cases with huge amount of thymic lymphoid cells attract other immune cell types in higher quantities.
Biological and Clinical Features in Relation With PD-L1 Expression in TETs
There was no significant connection between the percentage of PD-L1 positive tumor cells (TC value) neither the percentage of PD-L1 positive immune cell-occupied tumor area (IC value) and the sex or age of the patients (P-values of 0.4887 and 0.4765 using χ2 or Fisher exact test) (Table 1). The size of the tumor did not influence significantly the TC and the IC values (P-values of 0.2464 and 0.6795 using χ2 or Fisher exact test). Tumors with higher Masaoka-Koga stage was associated with lower TC value (P=0.0494), whereas there was no difference in IC value (P=0.7766), and the combination of TC and IC values (TC<50% and IC<10% vs. TC≥50% or IC≥10%; P=0.6442, using χ2 test) (Table 1.).
Cases with thymoma showed significantly longer overall survival compared with thymic carcinomas using Log-rank analysis (P=0.0154; mean 40.55 mo vs. 23.93 mo, respectively). Both thymomas and thymic carcinomas were divided into 2 groups based on PD-L1 expression in tumor cells and immune cells (TC<50% and IC<10% vs. TC ≥50% or IC ≥10%). Although a virtually negative effect on overall survival (OS) was observed in TETs expressing PD-L1, there was no statistically significant difference in the OS between the above-mentioned subgroups (TC<50% and IC<10% vs. TC ≥50% or IC ≥10%) in thymomas (P=0.1761) or in thymic carcinomas (P=0.5637) (Fig. 4).
Functional mutations of EGFR exon 19 and exon 21, KRAS exon 2, BRAF exon 15 analyzed by DNA-sequencing could not be detected (wild type) in any of the cases with thymic carcinoma.
PD1 Expression in Thymic Epithelial Tumors
PD1 expression could be evaluated in 35 of 36 cases as one case showed dubious positivity in nonrelated cells (granulocytes). PD-1 IHC staining was completely negative in TET tumor cells in all of our cases. In TETs, the dominant fraction of PD1 positive immune cells generally accumulated in tumor cell free zones of the tumor area (stromal area enclosed by tumor nests), and they were relatively spare among tumor cells sheaths (Fig. 5A). PD1 positive immune cells tended to form small clusters especially around small vessels and primary or secondary lymphoid follicles (Fig. 5B).
For PD1 expression the percentage of tumor area occupied by PD1 positive immune cells was determined. This value was <10% in all cases. In total, 7 of 29 (24.1%) thymomas and 3 of 7 (42.8%) thymic carcinomas were negative (<1%), 14 of 29 (48.3%) thymomas showed low-expression (1% to 5%), 8 of 29 (27.6%) thymomas, and 3 of 7 (42.8%) thymic carcinomas showed high-expression (≥5%). One case with thymic carcinoma presented with nonspecific positivity. There was no significant difference between thymomas and thymic carcinomas regarding PD1 expression rate (P=0.8898; mean 2.50 vs. 4.17, respectively). A significant correlation between the percentage of area occupied by PD1 and PD-L1 positive immune cells could be stated (P=0.0035; Spearman r=0.4934) when all cases were evaluated. The percentage of area occupied by PD1 positive immune cells did not show correlation with the percentage of PD-L1 positive tumor cells (TC value) when we analyzed all cases (P=0.1646; Spearman r=−0.2439). Focusing only on PD1 positive cases, TC and IC values showed very weak inverse correlation which was not significant (P=0.6970; Spearman r=−0.0819).
PD-L1 is an immune modulator whose interaction with the receptor PD-1 promotes adaptive resistance. Activation of the PD-1/PD-L1 pathway results in the decrease of antitumor immune response by the inhibition of T-cell receptor signaling. PD-L1 can be expressed both by tumor cells and tumor infiltrating immune cells. The therapeutic inhibition of this pathway is a recently evolved promising approach in different solid tumors such as NSCLC, melanoma, or renal cell carcinoma. However, the proper representation of the PD-1/PD-L1 complex within the tumor is individual and seems to have an important role influencing the response rate.11 Several clinical studies used PD-L1 immunohistochemistry for stratification and found that reliable determination of the target PD-L1 is predictive of treatment response and recommend IHC as a companion diagnostic tool to support treatment decision in specific histologic entities.13
Immune checkpoint inhibitors are promising targeted approaches, but little is known about the efficiency of this therapy in thymic epithelial tumors because of the lack of randomized clinical studies. Only one single institution phase II study is available, which was designed to check the clinical effect (response rate, progression free survival, and overall survival) of pembrolizumab in patients with recurrent thymic carcinoma. This study has analyzed 40 patients with a response rate of 22.5%, the median progression free survival was 4.2 months, and the median overall survival was 24.9 months. These results suggest survival benefit compared with the conventional chemotherapy. In contrast, this study has found higher frequency of immune-related adverse events in thymic carcinomas (6/40) comparing with other tumors treated with pembrolizumab.17
Unfortunately, a number of PD-L1 antibody clones are available for histology and they require different protocols, scoring, and cut-off values. The expression pattern in tumor cells and immune cells can be different as well. Because of these variables the predictive value of the IHC analysis depends on the clone of PD-L1 antibody, the histologic subtype of the primary tumor and the specific type of PD1/PD-L1 inhibitor as well. Although there is a growing number of publications regarding major cancer types, there is limited information focusing on immune checkpoint status of rare malignancies, such as thymic epithelial tumors. According to studies presented so far, TETs generally express PD-L1 protein, but evaluation and presentation were not done using presently accepted protocols.14,15
Katsuya and colleagues have analyzed the PD-L1 expression of thymic epithelial tumors using PD-L1 antibody clone E1L3N. The tumor PD-L1 expression score was calculated semiquantitatively by multiplying intensity (0 to 3) by the staining area (0 to 100%). 70% of thymic carcinoma (26/37) and 23% of thymoma (22/102) samples stained positive for PD-L1. Padda and colleagues have used PD-L1 antibody clone 5H1 and the staining was scored based on the intensity. The score ranged between 0 and 3, where score 3 was categorized as PD-L1high. PD-L1high staining was observed in 68.1% (47/69) of thymic epithelial tumor cases involved in the study.14,15
In our study, the procedure accepted for the evaluation of NSCLC samples was used that was approved as complementary diagnostic test predicting atezolizumab efficacy.12 In line with this, the Ventana PD-L1 clone SP142 primary antibody was used for the evaluation of PD-L1 expression. Altogether 36 thymic epithelial tumors were analyzed that was possible with only little adaptation of the original evaluation scheme considering the specialties of TET histomorphology. A major difference to consider seemed the nature of immune cells present in the samples. The thymus as a lymphatic organ is basically composed of thymocytes of T-cell origin that may be present in larger rests surrounding neoplastic epithelial proliferations. According to our interpretation, only intratumor immune cells were considered that were placed in an immediate proximity to the epithelial cell compartment.
Tumor cells and tumor infiltrating immune cells were determined separately. TC value was the percentage of positive tumor cells and IC value was the percentage of tumor area occupied by positive immune cells as described before using this specific PD-L1 clone. TET samples with less than 1% positivity for TC or IC were considered as IHC negative. Positive cases were categorized by a cut-off value introduced originally for NSCLCs.16
The method was well applicable with good reproducibility. In certain subtypes of TETs such as type B1 thymoma, the percentage of lymphoid cells reached over 80% of the area making the exact identification of the epithelial component difficult. In these cases, an additional pan-cytokeratin IHC staining was applied to highlight the epithelial component.
We stated that the IHC determination of PD-L1 proposed for lung carcinoma is well applicable and is a method of choice for treatment selection in thymic neoplasia. A high percentage of TETs show PD-L1 expression in tumor cells and also in tumor infiltrating immune cells. Combining TC and IC values 69% of thymomas (20/29) and 43% of thymic carcinomas (3/7) reached the cut-off values applied (TC≥50% or IC≥10%). There was no statistically significant difference in the percentage of tumor cell PD-L1 positivity in thymoma and thymic carcinoma subtypes. Further, TC or IC positivity did not correlate with any clinical-pathologic variable in TET. The individual variability in PD-L1 expression especially in thymic carcinomas may have a clinical impact reflecting different response potential when inhibitors are applied.
There was no significant difference in the overall survival between PD-L1 high expressor (TC≥50% or IC≥10%) and negative/low expressor cases (TC<50% and IC<10%) neither in thymomas nor in thymic carcinomas. This result is consistent with a previous study where unadjusted analysis showed the same trend, but when adjusted for age and sex, cases with intermediate-strong PD-L1 expression had a significantly worse OS.14 Unfortunately, the use of a different antibody clone and scoring system does not allow further comparisons. The impact of PD-L1 expression on prognosis is still generally debated.14 PD-L1 expression was associated with favorable prognosis in a defined subgroup of lung cancer or colorectal cancer,18,19 whereas unfavorable outcome was observed in breast cancer or in a subgroup of Hodgkin-lymphoma.20,21 Beside biological differences, the data regarding PD-L1 expression and prognosis are challenged by technical variables. Several different antibody clones, staining protocols, and evaluation schemes were used even in the same tumor type.
The clinical role of PD1 receptor expression was also investigated in lymphoid and solid neoplasias by several studies.9–11 Similar to earlier reports in NSCLC where PD1 expression did not have prognostic relevance,19 epithelial cells seemed to be completely negative in the thymus irrespective of the histologic diagnosis. On the contrary, intratumor PD1+ immune cells were present at low cell numbers in all cases. PD1 expression was generally <10% and was not different between thymomas and thymic carcinomas. It is interesting to note that, the extent of the PD-1+ infiltrate showed a significant correlation with the percentage of PD-L1 positive immune cells (P=0.0035). This result contradicts with a previous study,15 where the PD1 expression appeared to be independent of the PD-L1 score. Again, the difference may be explained by the use of an alternative antibody clone and evaluation method.
TETs contain more immune cells than other solid neoplasms. These lymphocytes are permanent component of the particular TET subtypes, and they should not be considered as tumor infiltrating immune cells. In our study, there was significant correlation between the amount of lymphoid component and IC value (percentage of tumor area infiltrated by PD-L1 positive immune cells). On the basis of the distribution and cytomorphology, among immune cells, mainly macrophages and dendritic cells were found to be positive for PD-L1 in our material, whereas intratumor lymphocytes generally proved to be negative.
Thymic epithelial tumors are infrequent neoplasias the treatment of which rarely needs a systemic oncological approach. The application of individualized treatments in aggressive cases could be considered on the basis of potential biological targets. Immune checkpoint inhibitors would be ideal if requirements for the inhibition are provided. Our study demonstrated that PD-L1 IHC analysis with the previously accepted criteria in NSCLC is technically applicable in TETs and may be used for PD-1/PD-L1 treatment stratification.
1. Marx A, Chan KCJ, Coindre JM, et al. The 2015 World Health Organization classification of tumors of the thymus. J Thoracic Oncol. 2015;10:1383–1395.
2. Kondo K, Monden Y. Therapy for thymic epithelial tumors
: a clinical study of 1,320 patients from Japan. Ann Thorac Surg. 2003;76:878–885.
3. Jackson MA, Ball DL. Post-operative radiotherapy in invasive thymoma. Radiother Oncol. 1991;21:77–82.
4. Rashid OM, Cassano AD, Takabe K. Thymic neoplasm: a rare disease with a complex clinical presentation. J Thorac Dis. 2013;5:173–183.
5. Ströbel P, Hohenberger P, Marx A. Thymoma and thymic carcinoma: molecular pathology and targeted therapy. J Thoracic Oncol. 2010;5(10 suppl 4):S286–S290.
6. Merveilleux du Vignaux C, Maury JM, Girard N. Novel agents in the treatment of thymic malignancies. Curr Treat Options Oncol. 2017;18:52.
7. Song Z, Yu X, Zhang Y. Rare frequency of gene variation and survival analysis in thymic epithelial tumors
. Onco Targets Ther. 2016;9:6337–6342.
8. Girard N, Shen R, Guo T, et al. Comprehensive genomic analysis reveals clinically relevant molecular distinctions between thymic carcinomas and thymomas. Clin Cancer Res. 2009;15:6790–6799.
9. Taube JM, Alison K, Brahmer JR, et al. Association of PD-1, PD-1 ligands, and other features of the tumor immune microenvironment with response to anti-PD-1 therapy. Clin Cancer Res. 2014;20:5064–5074.
10. Xia Y, Jeffrey Medeiros L, Young KH. Signaling pathway and dysregulation of PD1 and its ligands in lymphoid malignancies. Biochim Biophys Acta. 2016;1865:58–71.
11. Chen L, Han X. Anti-PD-1/PD-L1 therapy of human cancer: past, present, and future. J Clin Invest. 2015;125:3384–3391.
12. Ventana Medical Systems. Media release. Roche announces FDA approval for VENTANA PD-L1 (SP142) Assay to support immunotherapy treatment decisions in lung cancer. Available at: http://www.ventana.com/roche-receives-fda-approval-for-pd-l1-assay-for-nsclc+
. Accessed October 27, 2016.
13. Ratcliffe MJ, Sharpe A, Midha A, et al. Agreement between programmed cell death ligand-1 diagnostic assays across multiple protein expression cutoffs in non-small cell lung cancer. Clin Cancer Res. 2017;23:3585–3591.
14. Padda SK, Reiss JW, Schwartz EJ, et al. Diffuse high intensity PD-L1 staining in thymic epithelial tumors
. J Thorac Oncol. 2015;10:500–508.
15. Katsuya Y, Fujita Y, Horinouchi H, et al. Immunohistochemical status of PD-L1 in thymoma and thymic carcinoma. Lung Cancer. 2015;88:154–159.
16. Rittmeyer A, Barlesi F, Waterkamp D, et al. Atezolizumab versus docetaxel in patients with previously treated non-small-cell lung cancer (OAK): a phase 3, open-label, multicentre randomised controlled trial. Lancet. 2017;389:255–265.
17. Giaccone G, Kim C, Thompson J, et al. Pembrolizumab in patients with thymic carcinoma: a single-arm, single-centre, phase 2 study. The Lancet. 2018;19:347–355.
18. Droeser RA, Hirt C, Viehl CT, et al. Clinical impact of programmed cell death ligand 1 expression in colorectal cancer. Eur J Cancer. 2013;49:2233–2242.
19. Schmidt LA, Kümmel A, Görlich D, et al. PD-1 and PD-L1 expression in NSCLC indicate a favorable prognosis in defined subgroups. PLoS One. 2015;10:e0136023.
20. Muenst S, Schaerli AR, Gao F, et al. Expression of programmed death ligand 1 (PD-L1) is associated with poor prognosis in human breast cancer. Breast Cancer Res Treat. 2014;146:15–24.
21. Menter T, Bodmer-Haecki A, Dirnhofer S, et al. Evaluation of the diagnostic and prognostic value of PDL1 expression in Hodgkin and B-cell lymphomas. Hum Pathol. 2016;54:17–24.