Purpose: Immunotherapy targeting the PD-1/PD-L1 pathway has changed the treatment landscape of non–small cell lung carcinoma (NSCLC). We demonstrated that HHLA2, a newly identified immune inhibitory molecule, was widely expressed in NSCLC. We now compared the expression and function of PD-L1 with alternative immune checkpoints, B7x and HHLA2.

Experimental Design: Expression was examined in tissue microarrays consisting of 392 resected NSCLC tumors. Effects of PD-L1, B7x, and HHLA2 on human T-cell proliferation and cytokine production were investigated.

Results: PD-L1 expression was identified in 25% and 31% of tumors in the discovery and validation cohorts and was associated with higher stage and lymph node involvement. The multivariate analysis showed that stage, TIL status, and lymph node involvement were independently associated with PD-L1 expression. B7x was expressed in 69% and 68%, whereas HHLA2 was positive in 61% and 64% of tumors in the two sets. The coexpression of PD-L1 with B7x or HHLA2 was infrequent, 6% and 3%. The majority (78%) of PD-L1–negative cases expressed B7x, HHLA2, or both. The triple-positive group had more TIL infiltration than the triple-negative group. B7x-Ig and HHLA2-Ig inhibited TCR-mediated proliferation of CD4 and CD8 T cells more robustly than PD-L1-Ig. All three significantly suppressed cytokine productions by T cells.

Conclusions: The majority of PD-L1–negative lung cancers express alternative immune checkpoints. The roles of the B7x and HHLA2 pathway in mediating immune evasion in PD-L1–negative tumors deserve to be explored to provide the rationale for an effective immunotherapy strategy in these tumors. Clin Cancer Res; 24(8); 1954–64. ©2018 AACR.

This article is featured in Highlights of This Issue, p. 1771

Translational Relevance

Immunotherapy with antibodies against PD-1 or PD-L1 has shifted the treatment paradigm in lung cancer, despite modest efficacy as monotherapy in PD-L1–negative tumors. There is a strong interest in identifying new predictive biomarkers, resistance mechanisms, and new therapeutic targets. B7x and HHLA2 are newly discovered members of the B7-CD28 immune checkpoint family. Our study shows that both B7x and HHLA2 are widely expressed in lung cancers. There is limited coexpression of PD-L1 with B7x and HHLA2. Remarkably, B7x and HHLA2 are commonly found in PD-L1–negative tumors. Functionally, all three checkpoint molecules suppressed TCR-mediated proliferation of T cells and cytokine production from T cells. These results suggest that B7x and HHLA2 are potential immune escape mechanisms in these tumors and thus pave the way for designing novel combined or alternative immunotherapeutic strategies.

The B7 family of ligands and CD28 family of receptors play crucial roles in the regulation of T-cell responses by providing costimulatory and coinhibitory signals (1, 2). Immunotherapy with antibodies against the well-known elements of the family, such as PD-1, PD-L1, and CTLA4, has revolutionized the treatment paradigm for a variety of tumor types, including non–small cell lung cancer (NSCLC). The antibodies targeting the PD-1/PD-L1 pathway have demonstrated impressive clinical benefits with prolonged survival in comparison with conventional chemotherapy, resulting in FDA approval of nivolumab, pembrolizumab, and atezolizumab in NSCLC as second-line treatment, as well as pembrolizumab in treatment-naïve patients with tumor PD-L1 high lung cancers (3–6).

The tumor PD-L1 expression has been implicated as one of the predictive biomarkers for these agents. In the treatment of refractory NSCLC, high PD-L1 expression was associated with an higher response rates (around 30%–40%) from the PD-1/PD-L1inhibitors, whereas only 8% to 10% of patients with negative PD-L1 expression would respond to these antibodies (7). Nevertheless, PD-L1 expression is not a perfect biomarker, and several other factors have also been proposed in predicting the activities of the PD-1/PD-L1 inhibitors, for instance, immune evasion mechanisms involving alternative immune checkpoint molecules; immune activity surrogates like tumor-infiltrating immune cells [tumor-infiltrating lymphocytes (TIL)], T-cell receptor clonality and immune gene signatures; tumor foreignness such as mutational burden, neoantigen burden, and other surrogate markers (7–9).

B7x (B7-H4/B7S1) (10–12) and HHLA2 (B7H7/B7H5/B7y) (13–17) are recently discovered B7 family ligands, which potentially function as T-cell coinhibitory molecules as evidence suggests that they suppress proliferation and cytokine production of both CD4+ and CD8+ T cells. Given their emerging roles of suppressing tumor immune responses with likely different immune evasion mechanisms from PD-1/PD-L1 pathways (2, 18–24), there has been increasing interests to explore them as therapeutic immune targets (17, 25). We previously investigated the expression and clinical significance of the most recently discovered immune checkpoint, HHLA2, in human lung cancer and found that it is widely expressed in NSCLC across different subtypes and is associated with EGFR mutational status (26). The results suggest that HHLA2 may be a novel immune evasion mechanism within the lung cancer microenvironment and thus a potential target for lung cancer immunotherapy.

Despite their established expression in a significant fraction of human lung cancers, the expression association between HHLA2, B7x, and PD-L1 in any human cancer remains unknown. In the current study, we focused on determining the coexpression profiles and clinical significance of PD-L1, B7x, and HHLA2 in lung cancer, especially the contribution of B7x and HHLA2 in PD-L1–negative tumors. We also examined and compared their capability to suppress human CD4 and CD8 T-cell functions.

Patients and samples

As previously reported (26), the relevant clinical data and pathology information were collected through retrospective chart reviews. All protocols were reviewed and approved by the Institutional Review Board. Tissue microarrays (TMA) in the discovery cohort included 195 NSCLC tumor tissues, and the TMAs in the separate validation cohort consisted of 197 NSCLC cases (mostly resected stage I–III NSCLC specimens). The TMAs were prepared as 1.0 mm cores with all cases in triplicates and with at least one of the three cores at the periphery of the nodule/mass, but still containing tumor.

In the discovery cohort (n = 195 patients), the mean age of the patient population was 68.3 years. Among them with known staging information (n = 180), stage I, II, and III were 81%, 12%, and 7%, respectively. For patients in the validation cohort (n = 197), the mean age was 66.5 years. Stage I, II, and III were 70%, 16%, and 14%, respectively, in cases with staging details (n = 171). Histology wise, 64% of cases in the discovery group and 76% in the validation cohort were adenocarcinoma (Table 1).

Table 1.

Clinicopathologic characteristics of PD-L1 and B7x expression in patients with lung cancer

A. PD-L1
Discovery cohort (n = 195)Validation cohort (n = 197)
ParameterPD-L1 negativePD-L1 positivePParameterPD-L1 negativePD-L1 positiveP
Age, years 67.7 69.9 0.20 Age, years 65.2 69.1 0.15 
Gender   0.65 Gender   0.99 
 Female (n = 116) 86 (74%) 30 (26%)   Female (n = 108) 74 (69%) 34 (31%)  
 Male (n = 74) 57 (77%) 17 (23%)   Male (n = 73) 50 (68%) 23 (32%)  
Race   0.19 Race   0.57 
 Asian (n = 8) 6 (75%) 2 (25%)   Asian (n = 0)    
 AA (n = 15) 10 (67%) 5 (33%)   AA (n = 3) 3 (100%) 0 (0%)  
 Hispanic (n = 16) 12 (75%) 4 (25%)   Hispanic (n = 10) 6 (60%) 4 (40%)  
 White (n = 142) 111 (78%) 31 (22%)   White (n = 97) 61 (63%) 36 (37%)  
Histology   0.08 Histology   0.17 
 Adeno (n = 158) 124 (78%) 34 (22%)   Adeno (n = 149) 109 (73%) 40 (27%)  
 Squam (n = 8) 5 (63%) 3 (37%)   Squam (n = 23) 13 (57%) 10 (43%)  
 Large (n = 16) 9 (56%) 7 (44%)   Large (n = 2) 2 (100%) 0 (0%)  
Stage   0.003 Stage   0.006 
 I (n = 145) 117 (81%) 28 (19%)   I (n = 119) 89 (75%) 30 (25%)  
 II (n = 21) 12 (49%) 9 (21%)   II (n = 27) 19 (70%) 8 (30%)  
 III (n = 13) 6 (46%) 7 (54%)   III (n = 24) 10 (42%) 14 (58%)  
Lymph node involvement   0.04 Lymph node involvement   0.04 
 Neg (n = 144) 112 (78%) 32 (22%)   Neg (n = 124) 89 (72%) 35 (28%)  
 Pos (n = 27) 16 (59%) 11 (41%)   Pos (n = 37) 20 (54%) 17 (46%)  
Mutation status   0.07 Mutation status   0.94 
 EGFR (n = 19) 18 (95%) 1 (5%)   EGFR (n = 27) 19 (70%) 8 (30%)  
 KRAS (n = 20) 16 (80%) 4 (20%)   KRAS (n = 48) 33 (69%) 15 (31%)  
 WT/WT (n = 34) 23 (68%) 11 (32%)   WT/WT (n = 57) 38 (67%) 19 (33%)  
TIL score   0.14 TIL score   0.26 
 Absent (n = 10) 8 (80%) 2 (20%)   Absent (n = 3) 2 (67%) 1 (33%)  
 Low (n = 122) 96 (79%) 26 (21%)   Low (n = 117) 83 (71%) 34 (29%)  
 High (n = 54) 35 (65%) 19 (35%)   High (n = 57) 34 (60%) 23 (40%)  
B. B7x 
Discovery cohort (n = 195) Validation cohort (n = 197) 
Parameter B7x negative B7x positive P Parameter B7x negative B7x positive P 
Age, years 69.0 67.9 0.50 Age, years 68.2 66.1 0.43 
Gender   0.63 Gender   0.99 
 Female (n = 112) 36 (32%) 76 (68%)   Female (n = 103) 36 (35%) 67 (65%)  
 Male (n = 73) 21 (29%) 52 (71%)   Male (n = 70) 19 (27%) 51 (73%)  
Race   0.03 Race   0.85 
 Asian (n = 7) 1 (14%) 6 (86%)   Asian (n = 0)    
 AA (n = 14) 5 (36%) 9 (64%)   AA (n = 3) 1 (33%) 2 (67%)  
 Hispanic (n = 16) 8 (50%) 8 (50%)   Hispanic (n = 10) 5 (50%) 5 (50%)  
 White (n = 139) 37 (27%) 102 (73%)   White (n = 96) 35 (36%) 61 (64%)  
Histology   0.001 Histology   0.22 
 Adeno (n = 153) 37 (24%) 116 (76%)   Adeno (n = 138) 42 (30%) 96 (70%)  
 Squam (n = 8) 7 (87%) 1 (13%)   Squam (n = 24) 11 (46%) 13 (54%)  
 Large (n = 16) 8 (50%) 8 (50%)   Large (n = 2) 1 (50%) 1 (50%)  
Stage   0.71 Stage   0.08 
 I (n = 141) 42 (30%) 99 (70%)   I (n = 113) 42 (37%) 71 (63%)  
 II (n = 21) 8 (38%) 13 (62%)   II (n = 27) 4 (15%) 23 (85%)  
 III (n = 13) 4 (31%) 9 (69%)   III (n = 21) 7 (33%) 14 (67%)  
Lymph node (N per TNM staging)   0.94 Lymph node (N per TNM staging)   0.09 
 N0 (n = 140) 43 (31%) 97 (69%)   N0 (n = 122) 41 (34%) 81 (66%)  
 N1 (n = 16) 4 (25%) 12 (75%)   N1 (n = 19) 2 (11%) 17 (89%)  
 N2 (n = 10) 3 (30%) 7 (70%)   N2 (n = 15) 6 (40%) 9 (60%)  
Mutation status   0.13 Mutation status   0.13 
 EGFR (n = 17) 2 (12%) 15 (88%)   EGFR (n = 25) 7 (28%) 18 (72%)  
 KRAS (n = 20) 8 (40%) 12 (60%)   KRAS (n = 46) 11 (24%) 35 (76%)  
 WT/WT (n = 34) 12 (35%) 22 (65%)   WT/WT (n = 52) 22 (42%) 30 (58%)  
TIL score   0.15 TIL score   0.44 
 Absent (n = 10) 5 (50%) 5 (50%)   Absent (n = 3) 0 (0%) 3 (100%)  
 Low (n = 122) 35 (29%) 87 (71%)   Low (n = 115) 38 (33%) 77 (67%)  
 High (n = 54) 17 (32%) 36 (68%)   High (n = 58) 17 (29%) 41 (71%)  
A. PD-L1
Discovery cohort (n = 195)Validation cohort (n = 197)
ParameterPD-L1 negativePD-L1 positivePParameterPD-L1 negativePD-L1 positiveP
Age, years 67.7 69.9 0.20 Age, years 65.2 69.1 0.15 
Gender   0.65 Gender   0.99 
 Female (n = 116) 86 (74%) 30 (26%)   Female (n = 108) 74 (69%) 34 (31%)  
 Male (n = 74) 57 (77%) 17 (23%)   Male (n = 73) 50 (68%) 23 (32%)  
Race   0.19 Race   0.57 
 Asian (n = 8) 6 (75%) 2 (25%)   Asian (n = 0)    
 AA (n = 15) 10 (67%) 5 (33%)   AA (n = 3) 3 (100%) 0 (0%)  
 Hispanic (n = 16) 12 (75%) 4 (25%)   Hispanic (n = 10) 6 (60%) 4 (40%)  
 White (n = 142) 111 (78%) 31 (22%)   White (n = 97) 61 (63%) 36 (37%)  
Histology   0.08 Histology   0.17 
 Adeno (n = 158) 124 (78%) 34 (22%)   Adeno (n = 149) 109 (73%) 40 (27%)  
 Squam (n = 8) 5 (63%) 3 (37%)   Squam (n = 23) 13 (57%) 10 (43%)  
 Large (n = 16) 9 (56%) 7 (44%)   Large (n = 2) 2 (100%) 0 (0%)  
Stage   0.003 Stage   0.006 
 I (n = 145) 117 (81%) 28 (19%)   I (n = 119) 89 (75%) 30 (25%)  
 II (n = 21) 12 (49%) 9 (21%)   II (n = 27) 19 (70%) 8 (30%)  
 III (n = 13) 6 (46%) 7 (54%)   III (n = 24) 10 (42%) 14 (58%)  
Lymph node involvement   0.04 Lymph node involvement   0.04 
 Neg (n = 144) 112 (78%) 32 (22%)   Neg (n = 124) 89 (72%) 35 (28%)  
 Pos (n = 27) 16 (59%) 11 (41%)   Pos (n = 37) 20 (54%) 17 (46%)  
Mutation status   0.07 Mutation status   0.94 
 EGFR (n = 19) 18 (95%) 1 (5%)   EGFR (n = 27) 19 (70%) 8 (30%)  
 KRAS (n = 20) 16 (80%) 4 (20%)   KRAS (n = 48) 33 (69%) 15 (31%)  
 WT/WT (n = 34) 23 (68%) 11 (32%)   WT/WT (n = 57) 38 (67%) 19 (33%)  
TIL score   0.14 TIL score   0.26 
 Absent (n = 10) 8 (80%) 2 (20%)   Absent (n = 3) 2 (67%) 1 (33%)  
 Low (n = 122) 96 (79%) 26 (21%)   Low (n = 117) 83 (71%) 34 (29%)  
 High (n = 54) 35 (65%) 19 (35%)   High (n = 57) 34 (60%) 23 (40%)  
B. B7x 
Discovery cohort (n = 195) Validation cohort (n = 197) 
Parameter B7x negative B7x positive P Parameter B7x negative B7x positive P 
Age, years 69.0 67.9 0.50 Age, years 68.2 66.1 0.43 
Gender   0.63 Gender   0.99 
 Female (n = 112) 36 (32%) 76 (68%)   Female (n = 103) 36 (35%) 67 (65%)  
 Male (n = 73) 21 (29%) 52 (71%)   Male (n = 70) 19 (27%) 51 (73%)  
Race   0.03 Race   0.85 
 Asian (n = 7) 1 (14%) 6 (86%)   Asian (n = 0)    
 AA (n = 14) 5 (36%) 9 (64%)   AA (n = 3) 1 (33%) 2 (67%)  
 Hispanic (n = 16) 8 (50%) 8 (50%)   Hispanic (n = 10) 5 (50%) 5 (50%)  
 White (n = 139) 37 (27%) 102 (73%)   White (n = 96) 35 (36%) 61 (64%)  
Histology   0.001 Histology   0.22 
 Adeno (n = 153) 37 (24%) 116 (76%)   Adeno (n = 138) 42 (30%) 96 (70%)  
 Squam (n = 8) 7 (87%) 1 (13%)   Squam (n = 24) 11 (46%) 13 (54%)  
 Large (n = 16) 8 (50%) 8 (50%)   Large (n = 2) 1 (50%) 1 (50%)  
Stage   0.71 Stage   0.08 
 I (n = 141) 42 (30%) 99 (70%)   I (n = 113) 42 (37%) 71 (63%)  
 II (n = 21) 8 (38%) 13 (62%)   II (n = 27) 4 (15%) 23 (85%)  
 III (n = 13) 4 (31%) 9 (69%)   III (n = 21) 7 (33%) 14 (67%)  
Lymph node (N per TNM staging)   0.94 Lymph node (N per TNM staging)   0.09 
 N0 (n = 140) 43 (31%) 97 (69%)   N0 (n = 122) 41 (34%) 81 (66%)  
 N1 (n = 16) 4 (25%) 12 (75%)   N1 (n = 19) 2 (11%) 17 (89%)  
 N2 (n = 10) 3 (30%) 7 (70%)   N2 (n = 15) 6 (40%) 9 (60%)  
Mutation status   0.13 Mutation status   0.13 
 EGFR (n = 17) 2 (12%) 15 (88%)   EGFR (n = 25) 7 (28%) 18 (72%)  
 KRAS (n = 20) 8 (40%) 12 (60%)   KRAS (n = 46) 11 (24%) 35 (76%)  
 WT/WT (n = 34) 12 (35%) 22 (65%)   WT/WT (n = 52) 22 (42%) 30 (58%)  
TIL score   0.15 TIL score   0.44 
 Absent (n = 10) 5 (50%) 5 (50%)   Absent (n = 3) 0 (0%) 3 (100%)  
 Low (n = 122) 35 (29%) 87 (71%)   Low (n = 115) 38 (33%) 77 (67%)  
 High (n = 54) 17 (32%) 36 (68%)   High (n = 58) 17 (29%) 41 (71%)  

Abbreviation: AA, African American.

IHC and assessment of TILs

As described previously (26), formalin-fixed primary lung tumor tissue sections were deparaffinized, followed by antigen retrieval treatment with sodium citrate (10 mmol/L, pH 6.0). Endogenous peroxidase activity was blocked by Dako peroxidase blocking reagent (Dako Corporation). A rabbit PD-L1/CD274 (SP142) mAb (Spring Bioscience) was used at a dilution of 1:100; a mouse anti-human HHLA2 mAb (clone 566.1, IgG1; refs. 13, 17, 26) was used at a concentration of 4 μg/mL (dilution of 1:500); a monoclonal anti-B7x (clone 1H3, IgG1; refs. 25, 27) antibody was used at dilution of 1:500 for overnight incubation. The B7x and HHLA2 antibodies have been validated by previously published studies (25, 26, 28, 29). Then, Dako Envision system-HRP (Dako Corporation) was used by adding biotinylated link universal and streptavidin-HRP, followed by DAB chromogen (Dako Corporation) and hematoxylin nuclear counterstaining. The expression of PD-L1, B7x, and HHLA2 was evaluated using IHC in FFPE patients' specimens, and scoring was performed by two independent investigators. On the basis of the intensity of staining, PD-L1, B7x, and HHLA2 expression was quantified as 0, 1, 2, and 3 for absent, low, moderate, and high expression. The H-score was determined as the percentage of staining (proportion score) multiplied by an ordinal value corresponding to the maximum intensity score in the specimen. The results of PD-L1 staining are also reported as a tumor proportion score, which is the percentage of tumor cells showing membrane staining of PD-L1.

As described previously (26), the scoring of TILs was performed in the same slides used for H&E staining and was read using a three-tiered scale based on the visual estimation of the proportion of lymphocytic infiltration in each histospot. A score of “absent” indicated absence of TILs, “low” was defined as 1% to 30%, and “high” was defined as >30%.

Human T-cell proliferation assay, cytokine analysis, antibodies, and flow cytometry

Human T cells were purified with Human T Lymphocyte Enrichment Set-DM (BD IMagTM) from 11 donors. All protocols were reviewed and approved by the Institutional Review Board. Purified T cells were labeled with CFSE (5 μmol/L; Sigma) in PBS for 10 minutes at room temperature, followed by wash with complete medium for at least 3 times. The 96-well round-bottom plates were precoated with anti-human CD3 (OKT3; BioLegend), control human IgG (hIgG) at 5 μg/mL, human PD-L1-Ig (hPD-L1-Ig) at 5 μg/mL, human B7x-Ig (hB7x-Ig) at 5 μg/mL, or human HHLA2-Ig (hHHLA2-Ig) at 5 μg/mL (R&D Systems) in PBS overnight at 4°C as reported previously (13). Plates were then washed with PBS 3 times and incubated with CFSE-labeled purified T cells in RPMI1640 medium (containing 10% FBS with 100 U/mL penicillin, 100 μg/mL streptomycin, 10 mmol/L HEPES, and 50 μmol/L 2-mercaptoethanol) for 4 days. T-cell proliferation was analyzed with anti-CD4 and anti-CD8 antibodies by flow cytometry.

Supernatants were collected 3 days after T-cell cultures. Human Th Cytokine Panel (13-plex; BioLegend) was used for measurement of human IL5, IL13, IL2, IL6, IL9, IL10, IFNγ, TNFα, IL17F, IL17A, IL4, IL21, and IL22 according to the manufacturer's instructions.

Purified T cells were stained with combination of APC anti-CD3, PE anti-CD4, PE/Cy7 anti-CD8, and isotype controls (BioLegend). Samples were acquired with LSR Yellow (BD Biosciences) and analyzed with FlowJo (Treestar).

Statistical analysis

χ2 test or Fisher exact test was appropriately used to compare the distribution of each categorical variables between PD-L1 expression–positive and negative groups. Two-sample proportion test was preformed to test whether the frequency of positive PD-L1 expression is greater in one mutational, histologic, or ethnic subgroup than another for pairwise comparisons. Wilcoxon rank sum test was used to test whether the PD-L1 H-score was greater in one mutational, histologic, ethnic, or TIL subgroup than another. Logistic regression models were used for multivariate analysis. Comparison of the Kaplan–Meier survival curves between expression-positive and negative groups or different degree of TIL groups were performed using the log-rank test. Similar statistical approaches were used to analyze the expression profile of B7x. All tests were two-sided using the SAS 9.2 software (SAS Institute Inc.). All P values ≤0.05 were considered statistically significant.

Clinicopathologic characteristics of PD-L1 expression in patients with lung cancers

The clinicopathologic features of PD-L1 expression were first analyzed in the discovery cohort (n = 195 patients) and then reassessed in the separate validation set (n = 197 patients; Table 1A). As reported previously (26), the majority of patients (85%, 334/392) had resected stage I to III NSCLC. Mutational status was available for approximately half of the population (49%, 191/392), and histology was documented for the whole cohort. PD-L1 was expressed in 25% of the tumors in the discovery cohort and 31% of cases in the validation set (Table 1A). Positive PD-L1 cases were defined as samples with percentage of tumor cells revealing membranous staining of PD-L1 ≥ 1% (with intensity score 1, 2, or 3; Fig. 1A). The relationship between tumor PD-L1 expression and clinicopathologic variables in lung cancer were subsequently investigated (Table 1A). Univariate analysis revealed no significant correlation between PD-L1 expression and age, gender, race, histology, mutational status, and TIL scores in either discovery or validation cohort.

Figure 1.

A, Representative specimens demonstrating varying PD-L1 expression in lung cancers by IHC. B, Representative specimens demonstrating a range of B7x expression in lung cancer by IHC. The level of expression was graded as intensity of IHC staining: 0, absent staining; 1, weak staining; 2, moderate staining; 3, strong staining.

Figure 1.

A, Representative specimens demonstrating varying PD-L1 expression in lung cancers by IHC. B, Representative specimens demonstrating a range of B7x expression in lung cancer by IHC. The level of expression was graded as intensity of IHC staining: 0, absent staining; 1, weak staining; 2, moderate staining; 3, strong staining.

Close modal

PD-L1 expression by stage and lymph node status

In the discovery cohort, 19% (28/145) of stage I, 21% (9/21) of stage II, and 54% (7/13) of stage III lung cancers expressed PD-L1, and tumors with higher stage had higher PD-L1 expression (P = 0.003). Differential PD-L1 expression by stage (P = 0.006) was also identified in the validation cohort: 25% (30/119) of stage I, 30% (8/27) of stage II, and 58% (14/54) of stage III NSCLC cases were positive for PD-L1 (Table 1A). Pairwise comparison showed that the PD-L1 expression was significantly higher in stage III than stage I disease in both discovery and validation cohorts (Fig. 2).

Figure 2.

Proportion of PD-L1–positive cases and pairwise comparison in different stage or lymph node (LN) involvement groups (positive +, negative −) in the discovery cohort (A) and validation cohort (B). *, P < 0.05.

Figure 2.

Proportion of PD-L1–positive cases and pairwise comparison in different stage or lymph node (LN) involvement groups (positive +, negative −) in the discovery cohort (A) and validation cohort (B). *, P < 0.05.

Close modal

Relationship between PD-L1 expression and lymph node status was also determined. In the discovery set, 22% (32/144) of cases without lymph node involvement and 41% cases with positive lymph nodes expressed PD-L1 (P = 0.04), respectively. Similarly, 28% (35/124) of lymph node–negative cases and 46% lymph node–positive patients were found to have PD-L1 expression (P = 0.04) in the validation cohort (Table 1A). The correlation was further confirmed in the pairwise comparison (Fig. 2). Taken together, it appears that the tumor PD-L1 expression is associated with more aggressiveness, such as advanced stage and positive lymph node involvement (as expected given nodal status defining stage).

B7x expression in patients with lung cancers

B7x expression was identified in 69% (128/185) of cases in the discovery cohort and 68% (118/173) of patients in the validation set (Table 1B). Positive B7x cases were defined as samples with positive membranous and cytoplasmic B7x staining at intensity score 1, 2, or 3 (Fig. 1B). The B7x expression was not significantly different by age, gender, stage, lymph node involvement, mutational status, and TIL status. There was a significant difference in B7x expression between the racial groups in the discovery set (P = 0.03): 50% (8/16) of Hispanics, 64% (9/14) of African Americans, 73% (102/139) of non-Hispanic White, and 86% (6/7) of Asian patients had B7x-positive lung cancers. However, no significant difference existed in the validation cohort (P = 0.85): 50% (5/10) of Hispanics, 67% (2/3) of African Americans, and 64% (61/96) of non-Hispanic White patients expressed B7x. Similarly, significant difference in the B7x expression among distinct histology groups was only observed in the discovery set (P = 0.001) but not in the validation cohort (P = 0.22). Collectively, B7x is widely expressed in lung cancer across all the subgroups.

Comparison of H-scores for different groups

To further clarify the expression profiles of PD-L1 and B7x among different categories, H-scores were calculated between the groups. As shown in Table 2A, the PD-L1 H-scores of the more advanced stage III lung cancers were significantly higher than that of early stages (I and II) in both discovery cohort (P = 0.0004) and validation set (P = 0.0003). Similarly, lymph node–positive lung cancers had significantly higher PD-L1 H-scores than tumors without lymph node involvement in both discovery (P = 0.04) and validation sets (P = 0.03). On the other hand, significant difference between EGFR-mutated and WT/WT groups was only observed in the discovery (P = 0.03) cohort but not in the validation set (P = 0.95).

Table 2.

H-scores for PD-L1, and B7x expression correlated with stage, lymph nodes, and mutational status

A. PD-L1 expression:
Discovery cohortPValidation cohortP
Stage Early (I + II) Advanced (III) 0.004 Stage Early (I + II) Advanced (III) 0.0003 
Mean H-score (SD) 13 (40) 49 (64)  Mean H-score (SD) 21 (55) 64 (70)  
Lymph node Negative Positive 0.04 Lymph node Negative Positive 0.03 
Mean H-score (SD) 14 (42) 25 (49)  Mean H-score (SD) 23 (58) 40 (57)  
Mutation EGFR WT/WT 0.03 Mutation EGFR WT/WT 0.95 
Mean H-score (SD) 4 (17) 29 (59)  Mean H-score (SD) 27 (46) 23 (60)  
B. B7x expression: 
Discovery cohort   P Validation cohort   P 
Race Hispanic Non-Hispanic 0.1 Race Hispanic Non-Hispanic 0.36 
Mean H-score (SD) 51 (62) 85 (78)  Mean H-score (SD) 48 (58) 72 (72)  
Mutation EGFR WT/WT 0.06 Mutation EGFR WT/WT 0.55 
Mean H-score (SD) 108 (79) 66 (72)  Mean H-score (SD) 71 (59) 66 (69)  
Histology Adeno Non-Adeno 0.0002 Histology Adeno Non-Adeno 0.07 
Mean H-score (SD) 91 (77) (SQ + large) 34 (53)  Mean H-score (SD) 82 (72) (SQ + large) 55 (62)  
A. PD-L1 expression:
Discovery cohortPValidation cohortP
Stage Early (I + II) Advanced (III) 0.004 Stage Early (I + II) Advanced (III) 0.0003 
Mean H-score (SD) 13 (40) 49 (64)  Mean H-score (SD) 21 (55) 64 (70)  
Lymph node Negative Positive 0.04 Lymph node Negative Positive 0.03 
Mean H-score (SD) 14 (42) 25 (49)  Mean H-score (SD) 23 (58) 40 (57)  
Mutation EGFR WT/WT 0.03 Mutation EGFR WT/WT 0.95 
Mean H-score (SD) 4 (17) 29 (59)  Mean H-score (SD) 27 (46) 23 (60)  
B. B7x expression: 
Discovery cohort   P Validation cohort   P 
Race Hispanic Non-Hispanic 0.1 Race Hispanic Non-Hispanic 0.36 
Mean H-score (SD) 51 (62) 85 (78)  Mean H-score (SD) 48 (58) 72 (72)  
Mutation EGFR WT/WT 0.06 Mutation EGFR WT/WT 0.55 
Mean H-score (SD) 108 (79) 66 (72)  Mean H-score (SD) 71 (59) 66 (69)  
Histology Adeno Non-Adeno 0.0002 Histology Adeno Non-Adeno 0.07 
Mean H-score (SD) 91 (77) (SQ + large) 34 (53)  Mean H-score (SD) 82 (72) (SQ + large) 55 (62)  

The univariate analysis revealed differential B7x expression by race or histology in one of the two cohorts (Table 1B). As illustrated in Table 2B, no significant difference was observed for B7x H-scores between Hispanic and non-Hispanic groups in both cohorts (P = 0.1 and P = 0.36, respectively). Likewise, there was no significant difference between EGFR-mutated specimens and wild-type (WT/WT) tumors in both the discovery (P = 0.06) and validation cohorts (P = 0.55). Moreover, variable B7x tumor H-score expression in different histology groups was only found in the discovery cohort (P = 0.0002) but not in the validation set (P = 0.07).

Stage, TIL status, and lymph node involvement were independently associated with PD-L1 expression in lung cancer

Multivariate analysis of the whole cohort for PD-L1 expression showed that stage [stage III vs. stage I and II; OR = 2.866; 95% confidence interval (CI), 1.134–7.244; P = 0.02], TIL score (high vs. low; OR = 2.139; 95% CI, 1.244–3.679; P = 0.006), and lymph node involvement (positive vs. negative) (OR = 2.343; 95% CI, 1.094–5.017; P = 0.02) were independently associated with PD-L1 expression. Multivariate analysis for B7X expression showed that histology (adenocarcinoma vs. nonadenocarcinoma; OR = 2.959; 95% CI, 1.584–5.525; P = 0.0007) but not the TIL status (high vs. low; OR = 0.984; 95% CI, 0.590–1.638; P = 0.949) was independently associated with B7x expression.

Comparison of expression characteristics, clinicopathologic features, and survival among different B7 family members: PD-L1, B7x, and HHLA2

We then compared the expression profiles and clinical outcomes of the three B7 family members. In line with our previous report (26), HHLA2 was positive in most of lung cancer specimens, including 61% of tumors in the discovery set and 64% in the validation cohort. Representatives of triple-positive (positive staining in all three markers), double-positive (positive staining in two markers), and triple-negative (negative staining in all three markers) were illustrated in Fig. 3A–C. The frequencies of triple-positive and triple-negative were 13% and 15%, respectively (Fig. 3D). The triple-positive cases were significantly more frequent in the more advanced stage III lung cancers than that of early stages (26% in stage III vs. 12% in stage I and II, P = 0.0156). At the same time, there was a trend toward higher prevalence of triple-negative cases in early stages than that in the more advanced stage III (17% in stage I and II vs. 6% in stage III, P = 0.0991). Among double positives, tumors positive for both B7x and HHLA2 are common (41%), in comparison with infrequent cases with PD-L1 and another marker (6% for PD-L1 and B7x; 3% for PD-L1 and HHLA2). More importantly, the majority of PD-L1–negative cases (78%, n = 197/251) expressed one or both alternative immune checkpoints (B7x, HHLA2, or both).

Figure 3.

A, Representative triple-positive cases demonstrating triple-positive staining of HHLA2, B7x, and PD-L1. B, Representative double-positive cases. C, Representative triple-negative cases. Of note, we tried to show the same area in the same spots on the comparison fields. D, Frequency of triple-positive, triple-negative, double-positive, and single-positive cases.

Figure 3.

A, Representative triple-positive cases demonstrating triple-positive staining of HHLA2, B7x, and PD-L1. B, Representative double-positive cases. C, Representative triple-negative cases. Of note, we tried to show the same area in the same spots on the comparison fields. D, Frequency of triple-positive, triple-negative, double-positive, and single-positive cases.

Close modal

Furthermore, there was a significant association between TIL status and biomarker groups: The triple-positive group had more TIL infiltration than the triple-negative group (51.1% vs. 27.5%, P = 0.01). Survival analysis did not show any significant difference in survival between the immune checkpoint groups.

Effects of PD-L1, B7x, and HHLA2 on TCR-mediated proliferation of human CD4 and CD8 T cells

After defining the expression characteristics of PD-L1, B7x, and HHLA2, we next determined their effects on human T-cell proliferations. Human T cells were purified from 11 different donors and were activated with plate-bound human CD3 mAb and control hIgG, hPD-L1-Ig, hHHLA2-Ig, or hB7x-Ig. In comparison with control hIgG, hPD-L1-Ig inhibited 15% of TCR-mediated proliferation of CD4 T cells (P < 0.05) but not CD8 T cells. On the other hand, hB7x-Ig reduced 39% of CD4 (P < 0.001) and 14% of CD8 (P < 0.01) T-cell proliferation, whereas hHHLA2-Ig diminished 43% of CD4 (P < 0.001) and 26% of CD8 (P < 0.01) T-cell growth (Fig. 4A). Moreover, at the same concentration, in comparison with hPD-L1-Ig, both hB7x-Ig and hHHLA2-Ig were associated with significantly stronger inhibition of TCR-mediated proliferation of both human CD4 (P < 0.001) and CD8 T cells (P < 0.05 for hB7x-Ig and P < 0.01 for hHHLA2-Ig; Fig. 4A).

Figure 4.

A, The inhibitory effects of PD-L1, hB7x, and HHLA2 on TCR-mediated CD4 and CD8 T-cell proliferation. Human T cells were purified from 11 donors. CFSE-labeled T cells were stimulated with a combination of plate-bound anti-CD3 and control human IgG (hIgG) at 5 μg/mL, human PD-L1-Ig (hPD-L1-Ig) at 5 μg/mL, human B7x-Ig (hB7x-Ig) at 5 μg/mL, or human HHLA2-Ig (hHHLA2-Ig) at 5 μg/mL for 3 days. T cells were then stained with anti-CD4 and anti-CD8 antibodies and analyzed by flow cytometry. B, Inhibition of PD-L1, hB7x, and HHLA2 on cytokine production from T cells. Purified T cells were stimulated as above. Supernatants were collected 3 days after T-cell cultures and were used for measurement of human IL5, IL13, IL2, IL6, IL9, IL10, IFNγ, TNFα, IL17F, IL17A, IL4, IL21, and IL22. n = 11. *, P < 0.05; **, P < 0.01; ***, P < 0.001. n.s., not significant.

Figure 4.

A, The inhibitory effects of PD-L1, hB7x, and HHLA2 on TCR-mediated CD4 and CD8 T-cell proliferation. Human T cells were purified from 11 donors. CFSE-labeled T cells were stimulated with a combination of plate-bound anti-CD3 and control human IgG (hIgG) at 5 μg/mL, human PD-L1-Ig (hPD-L1-Ig) at 5 μg/mL, human B7x-Ig (hB7x-Ig) at 5 μg/mL, or human HHLA2-Ig (hHHLA2-Ig) at 5 μg/mL for 3 days. T cells were then stained with anti-CD4 and anti-CD8 antibodies and analyzed by flow cytometry. B, Inhibition of PD-L1, hB7x, and HHLA2 on cytokine production from T cells. Purified T cells were stimulated as above. Supernatants were collected 3 days after T-cell cultures and were used for measurement of human IL5, IL13, IL2, IL6, IL9, IL10, IFNγ, TNFα, IL17F, IL17A, IL4, IL21, and IL22. n = 11. *, P < 0.05; **, P < 0.01; ***, P < 0.001. n.s., not significant.

Close modal

Effects of PD-L1, B7x, and HHLA2 on cytokine production from T cells

We then investigated the effects of PD-L1, B7x, and HHLA2 on cytokine production from T cells (Fig. 4B). Among the 13 T cell–derived cytokines examined, hPD-L1-Ig significantly reduced the production of 11 cytokines: IFNγ (51% reduction), TNFα (38% reduction), IL5 (30% reduction), IL13 (43% reduction), IL2 (47% reduction), IL9 (82% reduction), IL10 (61% reduction), IL17A (55% reduction), IL4 (35% reduction), IL21 (39% reduction), and IL22 (47% reduction). Similarly, hB7x-Ig significantly inhibited production of 10 cytokines: IFNγ (53% reduction), TNFα (53% reduction), IL13 (38% reduction), IL9 (76% reduction), IL10 (78% reduction), IL17A (75% reduction), IL17F (79% reduction), IL4 (53% reduction), IL21 (49% reduction), and IL22 (62% reduction). In accordance with a previous report (13), hHHLA2-Ig significantly decreased the production of 8 cytokines: IFNγ (50% reduction), TNFα (51% reduction), IL13 (40% reduction), IL9 (79% reduction), IL10 (76% reduction), IL17A (70% reduction), IL4 (42% reduction), and IL22 (48% reduction). Thus, consistent with their functional roles as immune checkpoint molecules, PD-L1, B7x, and HHLA2 were able to significantly suppress TCR signaling–induced T-cell cytokine production.

The members of B7 immune checkpoint molecules, PD-L1, B7x, and HHLA2, play an important role in immune regulation. We previously showed that HHLA2 was widely expressed in two thirds of NSCLC cases, and its expression was significantly associated with EGFR mutational status (26). In the current study, we further defined the expression profiles and clinicopathologic features of PD-L1 and B7x in human lung cancers. We found that PD-L1 expression was higher in locally advanced, stage III than the early-stage (I and II) disease, and more prevalent in the more aggressive diseases with lymph node involvement than the ones without metastases. We also compared the relative expression of the B7 immune checkpoint molecules in lung cancers. In addition to the findings consistent with previous reports on limited coexpression between PD-L1 and B7x (30, 31), we noticed that PD-L1 was rarely coexpressed with HHLA2, whereas the coexpression of B7x and HHLA2 was common. More strikingly, there was frequent expression of B7x and HHLA2 in the PD-L1–negative tumors. The overall PD-L1 expression frequency in our study is similar to reported data in resected lung cancers (32, 33). On the other hand, there are controversies about the correlation between PD-L1 expression and lung cancer stages (30, 32, 33). The use of different PD-L1 antibodies/platforms and variable sample sizes with distinct stage distributions may at least partly contribute to the discrepancy. Similar correlation between PD-L1 positivity and advanced stage was identified in a comprehensive analysis when the same PD-L1 clone SP142 antibody was used to determine PD-L1 expression in 499 patients with surgically resected primary lung cancers (33).

Functionally, all three of checkpoint molecules suppressed TCR-mediated proliferation of T cells and cytokine production from T cells. To our knowledge, this is the first comparison of the functions of PD-L1, B7x, and HHLA2 on human T cells. As expected, all three of them negatively regulated T-cell responses, including suppressed T-cell proliferation and reduced cytokine secretion. Under the same concentration, B7x-Ig and HHLA2-Ig demonstrated even more potent inhibition of T-cell growth than PD-L1-Ig did. Moreover, the limited overlap between the expression of PD-L1 versus B7x and HHLA2 indicates their potential nonredundant biological functions or distinct spatial and/or temporal contributions to immune evasion.

B7x protein in tumor cells has at least three forms: transmembrane protein whose function is obvious, intracellular protein, and soluble protein whose functions are unknown. We previously reported half of renal cell carcinoma patients have elevated soluble B7x in the blood that was associated with advanced tumor stage (34). In this study, we found B7x had some intracellular staining within tumor cells (Fig. 1); in the future, it would be interesting to examine whether the intracellular B7x is the source of soluble B7x.

Despite controversies, PD-L1 expression has been used as one of the predictive biomarkers to guide PD-1/PD-L1–directed immunotherapy in lung cancers (35). For instance, as monotherapy, pembrolizumab is only approved in patients with PD-L1–expressing NSCLC based on efficacy analysis (3, 35). The limited benefits of PD-1/PD-L1 inhibitors in the PD-L1–negative tumors highlight the importance of searching for alternative or combination treatment strategies. Our findings of the frequent expression of the alternative immune molecules, B7x and HHLA2, in PD-L1–negative tumors may shed some light in this direction. At the same time, novel immunotherapeutic approaches targeting B7x and HHLA2 are under active development and have shown promising results in early studies (25, 36, 37). The wide expression of B7x and HHLA2 in lung cancer implicates the therapeutic potential of targeting those immune markers.

There are limitations in our study. The use of TMA sections in a retrospective study may curb tumor representation or heterogeneity of the markers. Moreover, to examine the coexpression profiles, we attempted to investigate the same area in the same spots on the comparison fields for all three molecules. It was challenging to do so for some of the cases. To overcome the spatial limitation, studies with multiplexed quantitative immunofluorescence are ongoing. In addition, the survival data are only available for a proportion of patients, and thus, our study is underpowered for survival analysis. Future studies to determine the biological significance of B7x and HHLA2 in lung cancers are warranted to further elucidate their contributions to tumoral immune escape and resistance to PD-1/PD-L1 inhibitors, especially in the PD-L1–negative tumors.

X. Zang is an inventor on patent number US 9447186 B2, which covers the topic of anti-B7x cancer immunotherapy and an inventor on two pending patents that cover the topic of HHLA2-directed immunotherapy. No potential conflicts of interest were disclosed by the other authors.

Conception and design: H. Cheng, M. Janakiram, X. Zang

Development of methodology: H. Cheng, A. Assal

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): H. Cheng, A. Borczuk, X. Ren, A. Assal

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): H. Cheng, A. Borczuk, M. Janakiram, X. Ren, J. Lin, A. Assal, B. Halmos, R. Perez-Soler

Writing, review, and/or revision of the manuscript: H. Cheng, A. Borczuk, M. Janakiram, X. Ren, B. Halmos, R. Perez-Soler, X. Zang

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): H. Cheng

Study supervision: H. Cheng, X. Zang

The work was supported by NIH R01CA175495 (to X. Zang), NIH R01DK100525 (to X. Zang), and Department of Defense PC131008 (to X. Zang).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Greenwald
RJ
,
Freeman
GJ
,
Sharpe
AH
. 
The B7 family revisited
.
Annu Rev Immunol
2005
;
23
:
515
48
.
2.
Zang
X
,
Allison
JP
. 
The B7 family and cancer therapy: costimulation and coinhibition
.
Clin Cancer Res
2007
;
13
:
5271
9
.
3.
Reck
M
,
Rodriguez-Abreu
D
,
Robinson
AG
,
Hui
R
,
Csoszi
T
,
Fulop
A
, et al
Pembrolizumab versus chemotherapy for PD-L1-positive non-small-cell lung cancer
.
N Engl J Med
2016
;
375
:
1823
33
.
4.
Borghaei
H
,
Paz-Ares
L
,
Horn
L
,
Spigel
DR
,
Steins
M
,
Ready
NE
, et al
Nivolumab versus docetaxel in advanced nonsquamous non-small-cell lung cancer
.
N Engl J Med
2015
;
373
:
1627
39
.
5.
Garon
EB
,
Rizvi
NA
,
Hui
R
,
Leighl
N
,
Balmanoukian
AS
,
Eder
JP
, et al
Pembrolizumab for the treatment of non-small-cell lung cancer
.
N Engl J Med
2015
;
372
:
2018
28
.
6.
Rittmeyer
A
,
Barlesi
F
,
Waterkamp
D
,
Park
K
,
Ciardiello
F
,
von Pawel
J
, 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
65
.
7.
Sacher
AG
,
Gandhi
L
. 
Biomarkers for the clinical use of PD-1/PD-L1 inhibitors in non-small-cell lung cancer: a review
.
JAMA Oncol
2016
;
2
:
1217
22
.
8.
Dijkstra
KK
,
Voabil
P
,
Schumacher
TN
,
Voest
EE
. 
Genomics- and transcriptomics-based patient selection for cancer treatment with immune checkpoint inhibitors: a review
.
JAMA Oncol
2016
;
2
:
1490
5
.
9.
Gibney
GT
,
Weiner
LM
,
Atkins
MB
. 
Predictive biomarkers for checkpoint inhibitor-based immunotherapy
.
Lancet Oncol
2016
;
17
:
e542
51
.
10.
Zang
X
,
Loke
P
,
Kim
J
,
Murphy
K
,
Waitz
R
,
Allison
JP
. 
B7x: a widely expressed B7 family member that inhibits T cell activation
.
Proc Natl Acad Sci U S A
2003
;
100
:
10388
92
11.
Sica
GL
,
Choi
IH
,
Zhu
G
,
Tamada
K
,
Wang
SD
,
Tamura
H
, et al
B7-H4, a molecule of the B7 family, negatively regulates T cell immunity
.
Immunity
2003
;
18
:
849
61
.
12.
Prasad
DV
,
Richards
S
,
Mai
XM
,
Dong
C
. 
B7S1, a novel B7 family member that negatively regulates T cell activation
.
Immunity
2003
;
18
:
863
73
.
13.
Zhao
R
,
Chinai
JM
,
Buhl
S
,
Scandiuzzi
L
,
Ray
A
,
Jeon
H
, et al
HHLA2 is a member of the B7 family and inhibits human CD4 and CD8 T-cell function
.
Proc Natl Acad Sci U S A
2013
;
110
:
9879
84
.
14.
Zhu
Y
,
Yao
S
,
Iliopoulou
BP
,
Han
X
,
Augustine
MM
,
Xu
H
, et al
B7-H5 costimulates human T cells via CD28H
.
Nat Commun
2013
;
4
:
2043
.
15.
Flajnik
MF
,
Tlapakova
T
,
Criscitiello
MF
,
Krylov
V
,
Ohta
Y
. 
Evolution of the B7 family: co-evolution of B7H6 and NKp30, identification of a new B7 family member, B7H7, and of B7′s historical relationship with the MHC
.
Immunogenetics
2012
;
64
:
571
90
.
16.
Mager
DL
,
Hunter
DG
,
Schertzer
M
,
Freeman
JD
. 
Endogenous retroviruses provide the primary polyadenylation signal for two new human genes (HHLA2 and HHLA3)
.
Genomics
1999
;
59
:
255
63
.
17.
Janakiram
M
,
Chinai
JM
,
Zhao
A
,
Sparano
JA
,
Zang
X
. 
HHLA2 and TMIGD2: new immunotherapeutic targets of the B7 and CD28 families
.
Oncoimmunology
2015
;
4
,
e1026534
.
18.
Zou
W
,
Chen
L
. 
Inhibitory B7-family molecules in the tumour microenvironment
.
Nat Rev Immunol
2008
;
8
:
467
77
.
19.
Zang
X
,
Thompson
RH
,
Al-Ahmadie
HA
,
Serio
AM
,
Reuter
VE
,
Eastham
JA
, et al
B7-H3 and B7x are highly expressed in human prostate cancer and associated with disease spread and poor outcome
.
Proc Natl Acad Sci U S A
2007
;
104
:
19458
63
.
20.
Barach
YS
,
Lee
JS
,
Zang
X
. 
T cell coinhibition in prostate cancer: new immune evasion pathways and emerging therapeutics
.
Trends Mol Med
2011
;
17
:
47
55
.
21.
Hofmeyer
KA
,
Ray
A
,
Zang
X
. 
The contrasting role of B7-H3
.
Proc Natl Acad Sci U S A
2008
;
105
:
10277
8
.
22.
Jeon
H
,
Ohaegbulam
KC
,
Abadi
YM
,
Zang
X
. 
B7x and myeloid-derived suppressor cells in the tumor microenvironment: a tale of two cities
.
Oncoimmunology
2013
;
2
:
e24744
.
23.
Janakiram
M
,
Abadi
YM
,
Sparano
JA
,
Zang
X
. 
T cell coinhibition and immunotherapy in human breast cancer
.
Discov Med
2012
;
14
:
229
36
.
24.
Zang
X
,
Sullivan
PS
,
Soslow
RA
,
Waitz
R
,
Reuter
VE
,
Wilton
A
, et al
Tumor associated endothelial expression of B7-H3 predicts survival in ovarian carcinomas
.
Mod Pathol
2010
;
23
:
1104
12
.
25.
Jeon
H
,
Vigdorovich
V
,
Garrett-Thomson
SC
,
Janakiram
M
,
Ramagopal
UA
,
Abadi
YM
, et al
Structure and cancer immunotherapy of the B7 family member B7x
.
Cell Rep
2014
;
9
:
1089
98
.
26.
Cheng
H
,
Janakiram
M
,
Borczuk
A
,
Lin
J
,
Qiu
W
,
Liu
H
, et al
HHLA2, a new immune checkpoint member of the B7 family, is widely expressed in human lung cancer and associated with EGFR mutational status
.
Clin Cancer Res
2017
;
23
:
825
32
.
27.
Wei
J
,
Loke
P
,
Zang
X
,
Allison
JP
. 
Tissue-specific expression of B7x protects from CD4 T cell-mediated autoimmunity
.
J Exp Med
2011
;
208
:
1683
94
.
28.
Koirala
P
,
Roth
ME
,
Gill
J
,
Chinai
JM
,
Ewart
MR
,
Piperdi
S
, et al
HHLA2, a member of the B7 family, is expressed in human osteosarcoma and is associated with metastases and worse survival
.
Sci Rep
2016
;
6
:
31154
.
29.
Janakiram
M
,
Chinai
JM
,
Fineberg
S
,
Fiser
A
,
Montagna
C
,
Medavarapu
R
, et al
Expression, clinical significance, and receptor identification of the newest B7 family member HHLA2 protein
.
Clin Cancer Res
2015
;
21
:
2359
66
.
30.
Schalper
KA
,
Carvajal-Hausdorf
D
,
McLaughlin
J
,
Altan
M
,
Velcheti
V
,
Gaule
P
, et al
Differential expression and significance of PD-L1, IDO-1, and B7-H4 in human lung cancer
.
Clin Cancer Res
2017
;
23
:
370
8
.
31.
Altan
M
,
Pelekanou
V
,
Schalper
KA
,
Toki
M
,
Gaule
P
,
Syrigos
K
, et al
B7-H3 Expression in NSCLC and its association with B7-H4, PD-L1 and tumor-infiltrating lymphocytes
.
Clin Cancer Res
2017
;
23
:
5202
9
.
32.
Wu
S
,
Shi
X
,
Sun
J
,
Liu
Y
,
Luo
Y
,
Liang
Z
, et al
The significance of programmed cell death ligand 1 expression in resected lung adenocarcinoma
.
Oncotarget
2017
;
8
:
16421
9
.
33.
Takada
K
,
Toyokawa
G
,
Okamoto
T
,
Shimokawa
M
,
Kozuma
Y
,
Matsubara
T
, et al
A comprehensive analysis of programmed cell death ligand-1 expression with the clone SP142 antibody in non-small-cell lung cancer patients
.
Clin Lung Cancer
2017
;
18
:
572
82
.
34.
Thompson
RH
,
Zang
X
,
Lohse
CM
,
Leibovich
BC
,
Slovin
SF
,
Reuter
VE
, et al
Serum-soluble B7x is elevated in renal cell carcinoma patients and is associated with advanced stage
.
Cancer Res
2008
;
68
:
6054
8
.
35.
Reck
M
,
Rabe
KF
. 
Precision diagnosis and treatment for advanced non-small- cell lung cancer
.
N Engl J Med
2017
;
377
:
849
61
.
36.
Janakiram
M
,
Shah
UA
,
Liu
W
,
Zhao
A
,
Schoenberg
MP
,
Zang
X
. 
The third group of the B7-CD28 immune checkpoint family: HHLA2, TMIGD2, B7x, and B7-H3
.
Immunol Rev
2017
;
276
:
26
39
.
37.
Dangaj
D
,
Lanitis
E
,
Zhao
A
,
Joshi
S
,
Cheng
Y
,
Sandaltzopoulos
R
, et al
Novel recombinant human b7-h4 antibodies overcome tumoral immune escape to potentiate T- cell antitumor responses
.
Cancer Res
2013
;
73
:
4820
9
.