METHODS: Prostatectomy tissue from 2781 patients were profiled using the Affymetrix HuEx 1.0 ST microarray. Pairwise comparisons were used to identify significant associations between B7-H3 expression and clinicopathologic variables, and survival analyses were used to evaluate the prognostic significance of B7-H3. Pearson’s correlation analyses were performed to assess the relationship of B7-H3 expression with molecular subtypes and individual transcripts. Androgen receptor (AR) occupancy at the B7-H3 locus was determined using chromatin immunoprecipitation (ChIP), and androgen-dependent expression changes in B7-H3 was evaluated by quantitative reverse transcription PCR in LNCaP cell lines. Oncomine was queried to evaluate B7-H3 expression in metastatic disease.
RESULTS: B7-H3 mRNA expression was positively associated with higher Gleason score (Po0.001), tumor stage (Po0.001), and castrate-resistant metastatic disease (Po0.0001). High B7-H3 expression correlated with the development of metastasis and prostate cancer-specific mortality, but this was not significant on multivariable analysis. B7-H3 expression correlated with ERG-positive disease (r = 0.99) and AR expression (r = 0.36). ChIP revealed an AR-binding site upstream of B7-H3, and the presence of androgens decreased B7-H3 expression in LNCaP suggesting potential direct AR regulation. Gene set enrichment analysis demonstrated an association of B7-H3 with androgen signaling as well as immune regulatory pathways.
CONCLUSIONS: Higher B7-H3 expression correlates with Gleason grade, prostate cancer stage and poor oncologic outcomes in prostatectomy cohorts. B7-H3 expression appears to be related to androgen signaling as well as the immune reactome.
Prostate Cancer and Prostatic Diseases (2017) 20, 28–35; doi:10.1038/pcan.2016.49; published online 1 November 2016
INTRODUCTION: T-cell activation requires the engagement of the T-cell receptor but additionally engagement of co-stimulatory molecules, most notably CD28 that binds B7-1 and B7-2 on antigen presenting cells. Several molecules sharing homology to B7 have been identified and constitute a B7 superfamily.1 Among these molecules, B7-H1 (PD-L1) has been shown to have an important role as an immune checkpoint ligand within the tissue microenvironment and can be targeted by humanized antibodies to allow for anti-tumor responses in several malignancies including advanced melanoma, lung, bladder and renal cancers.2 B7-H3 (CD276) was identified from a human dendritic-cell-derived cDNA library and shares roughly 20–27% amino-acid identity with other B7 family members.3
B7-H3 is expressed in multiple tissue types, including the epithelial cells of tumors. The expression is also inducible on the surface of T cells, dendritic cell and monocytes.3 The receptor, regulation and mechanism of action of B7-H3 are not fully known, but recent preclinical studies suggest varied effects of B7-H3 depending on the mechanism of inflammation and involved T-cell subset.4 In their work, utilizing a B7-H3 knockout model and various modes of inflammation, Luo et al.4 determined that although B7-H3 may inhibit Th2 type responses, it may act as a co-stimulatory molecule for Th1 and Th17 type responses. Th1 responses serve to regulate CD8+ T cells and are thought to be critical for recognition of tumor antigens. In line with the findings above, the expression of B7-H3 in tumors promotes tumor regression and knockout of B7-H3 increases tumor growth in mouse models of spontaneous cancer.5–7
At the protein level, B7-H3 expression has been investigated in various malignancies, most notably prostate cancer. Utilizing immunohistochemistry, Roth et al.8 demonstrated that B7-H3 is expressed in normal prostate epithelium and is more intensely expressed by prostate cancer. Within their series, virtually all prostate cancer expressed B7-H3, however, the intensity of expression varied with cancers exhibiting more aggressive phenotypes (larger tumors, those with extraprostatic extension) expressing higher levels of the protein. In addition, elevated levels of B7-H3 expression correlated with disease progression following surgery. These results were supported by work from Zang et al.,9 which also demonstrated increased expression of B7-H3 in the majority of prostate cancers, particularly in tumors that were nonorgan confined at surgery. In their work, strongly positive B7-H3 expression was prognostic of clinical failure and death from prostate cancer in univariate models. Further correlating with these data, high B7-H3 expression was also found to be prognostic of disease progression following salvage radiation therapy for recurrent prostate cancer after surgery.10 In addition to these studies on the primary prostate cancer tissue, B7-H3 protein expression has also been explored in metastasis and in prostate cancer treated with androgen deprivation. Here B7-H3 expression was identified in prostate cancer bone metastasis with a trend towards increased expression upon androgen deprivation.11 In this work, neoadjuvant androgen deprivation did not appear to alter expression in prostatectomy tissue.
Here we studied B7-H3 expression at the transcript level in two large prostatectomy series. Further, we used genome-wide expression data to examine B7-H3 expression among different molecular subtypes of prostate cancer and to correlate its expression with immune regulatory pathways and with androgen receptor signaling.
MATERIALS AND METHODS
Prostatectomy tissue was derived from two patient cohorts. The first included prostatectomy samples with associated genomic information from 2,111 patients prospectively submitted for clinical Decipher testing.12 A second cohort included prostatectomy tissue from 670 patients that had undergone radical prostatectomies at the Johns Hopkins Medical Institute (JHMI). In the patients from JHMI, two case-cohort designs were used to investigate clinical outcomes: (1) a case-cohort based on 260 men with intermediate or high risk localized prostate cancer undergoing prostatectomy at JHMI and then followed expectantly until clinical metastasis,13 and (2) a case-cohort natural history study of 211 patients who had biochemical recurrence after prostatectomy but did not receive therapy until the time of metastasis.
Prostatectomy sample selection and processing
Specimen selection, RNA extraction and microarray hybridization was done in a Clinical Laboratory Improvement Amendments-certified laboratory facility (GenomeDx Biosciences, San Diego, CA, USA) as previously described.13,14 Briefly, total RNA was extracted and purified using the RNeasy FFPE kit (Qiagen, Valencia, CA, USA). RNA was amplified and labeled using the Ovation WTA FFPE system (NuGen, San Carlos, CA, USA) and hybridized to Human Exon 1.0 ST GeneChips (Affymetrix, Santa Clara, CA, USA).
Quality control was performed using Affymetrix Power Tools, and normalization was performed using the Single Channel Array Normalization algorithm.15 Gene expression was summarized using the Affymetrix
core transcript cluster and corrected for batch effects using an empirical Bayes framework.
Publically available datasets of AR chromatin immunoprecipitation (ChIP)-Seq experiments were analyzed using IGV.16,17 A putative androgen-induced AR-binding site was identified upstream of the B7-H3 (CD276) gene. Chromatin immunoprecipitation experiments were performed as described previously.18 In brief, formaldehyde cross-linked LNCaP cells were subjected to immunoprecipitation with AR specific antibodies (Millipore, Darmstadt, Germany) or control IgG (Cell Signaling Technologies, Danvers, MA, USA) after 8 h of 100 nM dihydrotestosterone (DHT, Sigma Aldrich, St. Louis, MO, USA) or solvent control treatment. Enriched libraries were amplified using primers specific to the putative upstream regulatory site (upstream, F: 5′-GCTTTTATGAGCCTCCGTGA-3′; R: 5′-AGCACTGAGCCATTCACCTT-3′) and the transcriptional start site (TSS, F: 5′-CGTCCCTGAGTCCCAGAGT-3′; R: 5′-GGTTCCCGGGACTCCTGT-3′). Data are shown as relative enrichment normalized to input DNA. Primers specific to the KLK3 locus (which encodes for PSA and harbors a well characterized AR-binding site) were used as a control.
Androgen-dependent expression of B7-H3 in prostate cancer cell lines
LNCaP cells were grown in RPMI and either charcoal-stripped fetal bovine serum (FBS) (deprived of androgens), FBS (contains androgens) or charcoal stripped FBS supplemented with DHT. Cells were harvested and B7-H3 expression was assayed using quantitative reverse transcription PCR with Taqman assays. All experiments were done in quadruplicate. LNCaP cells were obtained from ATCC, authenticity was validated by STR profiling and cells were confirmed as mycoplasma free by the MycoAlert Mycoplasma Detection Kit (Lonza, Basel, Switzerland).
B7-H3 expression distribution was compared with the expression of 3 other known checkpoint molecules. Survival analysis was performed in the JHMI cohorts to evaluate the prognostic value of B7-H3 and to compare the prognostic value of the 4 ligands. Kaplan–Meier curves were constructed with the appropriate weighting of the cohorts to account for the case-cohort study design.19 To test the association of B7-H3 expression with clinicopathologic variables pair-wise comparisons using Wilcoxon ranksum test and, where appropriate, adjusted P-values via Bonferroni correction. Associations of B7-H3 expression with molecular subtypes (ERG+, ERG-ETS+, ERG-SPINK1+ and TripleNeg) 20 were evaluated using Pearson’s correlation. The association of B7-H3 expression with AR was assessed using Pearson’s correlation. Statistical analyses were performed in R, and all statistical tests were two-sided using a 5% significance level.
Functional characterization and GSEA
Associations of B7-H3 expression with 104 genes which are a part of Nanostring’s PanCancer Immune Profiling Panel were evaluated using Pearson’s correlation and P-values were adjusted using Holm’s method. Genes were ranked based on their Pearson’s correlation to B7-H3 in the prospective cohort and gene set enrichment analysis (GSEA) was carried out to identify gene sets associated with B7-H3 by using the GSEA analysis tool downloaded from the Broad Institute website. The curated gene sets of the Molecular Signature Database (MSigDB) version 4.0 was used for enrichment. The false discovery rate for GSEA is the estimated probability that a gene set with a given NES (normalized enrichment score) represents a false-positive finding, and a false discovery rateo0.25 is considered to be statistically significant for GSEA. The Database for Annotation, Visualization and Integrated Discovery (DAVID) was used to evaluate molecular concepts enriched in genes which were highly correlated with B7-H3 (ref. 21).
Relationship of B7-H3 and Immune ligands to pathologic and clinical correlates
When compared with benign prostatic tissue, B7-H3 expression is significantly higher in prostate cancers (Po0.04; Supplementary Figure 1).22 In order to further examine B7-H3 expression in prostate cancer, we analyzed 2111 prostatectomy samples for which microarray data were available. Median B7-H3 expression was within the top 19th percentile of genes expressed in prostate cancer and was 2.5-fold higher than the median expression for all genes. By contrast, and concordant with previous studies, PD-L1, PD-L2 and B7-H4 were not highly expressed in prostate cancer (Figure 1a). Expression of B7-H3 varied significantly but modestly according to stage and grade, with median expression increasing among those men with primary Gleason pattern 4 or Gleason sum 8–10 disease or those with non-organ confined disease (Po0.001 for both; Figure 1b and Supplementary Table 1). B7-H3 expression was significantly increased in castrate-resistant metastatic prostate cancer when compared with localized disease (Figure 1c).
To determine the correlation between increased B7-H3 expression and oncologic outcomes following surgery we performed Kaplan–Meier analysis for metastatic events based on median split of B7-H3 expression in cohorts for which long-term follow-up was available. Among men who underwent prostatectomy and then had no further treatment until the time of metastasis, higher B7-H3 expression (but not PD-L1, PD-L2 or B7-H4 expression) was significantly associated with metastatic outcome (Figure 2a). We additionally analyzed a separate case-cohort study of men who underwent radical prostatectomy and all of which developed biochemical recurrence but had no further treatment until metastatic progression. Among these men, who are higher risk of metastatic progression, increased B7-H3 expression also significantly correlated with the development of clinical metastasis (Figure 2b). To determine whether B7-H3 expression is independently associated with outcome, we performed mutli-variable analysis including standard prognostic variables for each cohort.
These studies showed that B7-H3 is not an independent predictor of outcome, likely due to the positive correlation of B7-H3 expression levels with grade and stage (Supplementary Table 2).
Relationship of B7-H3 to androgen receptor signaling
We next investigated the relationship between B7-H3 expression and molecular subtypes of prostate cancer.20 B7-H3 expression was relatively elevated among patients with ERG+ disease and decrease among triple negative patients (Figure 3a). A positive correlation between ERG+ status and B7-H3 expression suggested a relationship between B7-H3 and androgen receptor signaling. In regards to the correlation of B7-H3 expression with AR expression itself, we also found a positive correlation (Figure 3b), with B7-H3 being among the most correlated genes with AR expression (top 15th percentile) and visa versa (top 21st percentile) (Figures 3c and d). A similar finding was demonstrated in the JHMI cohorts with Pearson correlation coefficients of 0.47 and AR falling in the top 6th percentile of genes correlated with B7-H3 (Supplementary Figure 2). In gene set enrichment analysis, the androgen receptor signaling pathway was positively correlated with B7-H3 expression (normalized enrichment score = 1.62 and 1.90 P = 0.05 and o0.001 for the prospective and JHMI cohorts, respectively) (Figure 4). We next investigated whether AR would show androgen-induced binding at the B7-H3 (CD276) locus. Mining publically available ChIP-Seq data sets, we identified an AR-binding site upstream of B7-H3 (ref. 16). To corroborate this finding we performed AR ChIP in LNCaP cells and evaluated AR binding to a putative B7-H3 upstream regulatory element and the B7-H3 promoter in the absence or presence of androgen. Androgen stimulation resulted in a robust induction of AR binding to the B7-H3 upstream region, with enrichments similar to other well-characterized AR-binding sites (KLK3/PSA enhancer) (Figure 5 and Supplementary Figure 3).
To address the functional consequence of AR binding upstream of B7-H3 we evaluated the expression of B7-H3 under different conditions in the LNCaP prostate cancer cell line. LNCaP cells grown in RPMI supplemented with FBS (which contains androgens) show low expression of B7-H3 compared with cells grown in medium supplemented with charcoal-stripped FBS (deprived of androgens. Conversely, the expression of B7-H3 was greatly reduced in LNCaP cells grown in charcoal-stripped FBS, which was supplemented with DHT. These findings, shown in Figure 5c, suggest that androgen receptor binding negatively regulates B7-H3 expression. This finding is further supported by in silico analyses of publically available datasets supporting the notion that B7-H3 expression is suppressed by AR (Supplementary Figure 4).
Given the relationship between B7-H3 and AR, we conducted an exploratory analysis using genes having a Pearson’s correlation coefficient of 40.5 between B7-H3 and AR and used DAVID to functionally cluster those genes. Enriched clusters included genes involved in cell cycle, cell differentiation, proteolysis, apoptosis, splicing and DNA repair, as well as those associated with the androgen receptor and WNT signaling (Supplementary Figure 5).
Gene set enrichment analysis and immune gene correlations with B7-H3 expression
In order to define putative relationships and functions for B7-H3, we performed an unbiased gene set enrichment analysis using data derived from the prostatectomy tissue of the 2111 patients in the prospective cohort. Selected top positively enriched gene sets are shown in Table 1 with a complete list in Supplementary Table 3. No significantly negatively correlated gene set was found. Because of B7-H3’s role in immune modulation we were particularly interested in the enrichment of immune pathways. Significantly enriched immune-related gene sets included transforming growth factor (TGF)-β signaling and IL2-STAT5 signaling pathways (NES 1.64 and 1.6 respectively, adj Po0.05 for both). To further explore the associations between B7-H3 and immune-related genes we investigated the correlation between B7-H3 gene expression and expression of genes within a previously described immune panel (Supplementary Table 4). Supporting the enrichment of the TGF-β signaling pathway, we saw positive correlations between B7-H3 expression and SMAD2, IL17RA and RORC (Figure 6).
Recent successes in the treatment of melanoma, bladder cancer, lung cancer and kidney cancer via immune checkpoint blockade has galvanized efforts to understand the interplay between tumors and the immune system and to develop novel therapeutic strategies in immuno-oncology.23–25 Although immunotherapies have demonstrated some efficacy in prostate cancer, to date, their benefit has been somewhat marginal.26,27 B7-H3, is an attractive target for further study in prostate cancer based on immunohistochemical data suggesting its increased expression in aggressive and castrate-resistant metastatic prostate cancer and its homology to PD-L1 suggesting a potential role in immune modulation.8–11 Indeed, recently reported phase I results with Enoblituzumab (MGA271), an Fc-optimized monoclonal antibody targeting B7-H3 shoed clinical activity in advanced malignancies including prostate cancer.28 Here we examined B7-H3 at the level of gene expression and performed exploratory analyses to generate hypotheses regarding its relationship to the androgen axis and immune regulatory pathways in prostate cancer.
We found that B7-H3 was expressed in the majority of patients with localized prostate cancer. This was in contrast to PD-L1, PD-L2, and B7-H4, all of which showed low levels of expression and parallels findings of B7-H3 and other B7 homologs studied at the protein level. Similar to what is found in immunohistochemical studies, B7-H3 expression was associated with higher grade and stage disease, castrate-resistant prostate cancer, and early disease progression after local treatment. Our data thus support the increased expression of B7-H3 in aggressive disease and suggests that mRNA expression levels of B7-H3 may potentially act as a good surrogate for protein level expression.
It is worth noting that a number of animal studies suggest that B7-H3 has a stimulatory role in driving anti-tumor immunity, such that, contrary to the findings presented here—increased B7-H3 expression might be expected to correlate with improved clinical outcome. Indeed, studies of forced expression in a lymphoma model,6 a mastocytoma model7 and a colorectal cancer model29 suggest that B7-H3 interacts with a stimulatory receptor on CD8 T cells and that this receptor may be TLT-2 (ref. 30). Data in certain human cancers are consistent with an anti-tumor role for B7-H3; in both gastric31 and pancreatic cancer32 expression is associated with improved survival. In other human cancers such as renal cell carcinoma33 and prostate cancer (ref. 8 and data are shown here), B7-H3 expression is associated with a poor outcome. The reasons for this apparent discrepancy are not immediately obvious, but could potentially be explained by differential expression of activating versus inhibitory B7-H3 receptors on immune cells, or by differential infiltration of tumor subtypes with sets of immune cells. For example, expression of a stimulatory B7-H3 receptor on pro-tumor M2 macrophages or myeloid derived suppressor cells could occur in tumors in which B7-H3 expression is associated with poor outcome—here the stimulation of suppressive cells could lead to tumor progression. Conversely, in other tumor types the B7-H3 receptor could be expressed predominantly on tumorspecific CD8 T cells, so that receptor engagement could lead to an anti-tumor effect, as has been observed in most animal studies. This dual effect is not specific to B7-H3, infiltration of tumors with regulatory T cells is associated with poor outcome in most tumors,34 but not in colorectal cancer, where it is associated with an improved outcome.35 Identification of the human B7-H3 receptor(s) would provide crucial insights in terms of resolving these issues.
B7-H3 expression was strongly correlated with ERG positivity and negatively correlated with triple-negative tumors (ERG negative, ETS negative and SPNIK negative). This suggested a putative relationship between B7-H3 and androgen signaling, which was reinforced by gene set enrichment of androgen receptor signaling. In addition, a possible androgen response element was identified in the 5′ UTR of B7-H3, with occupancy being dependent on the presence of androgens. Taken together, these data suggest that B7-H3 expression may be regulated by the androgen signaling axis and indeed, in LNCaP cells we observed a suppressive effect of androgens on B7-H3 expression.
Whether B7-H3 has a proinflammatory or immunosuppressive role in human prostate cancer remains unclear. An examination of B7-H3 expression in relation to other immune pathways and genes suggested an association with TGF-β signaling. Further, a positive correlation with RORC and IL17 may be suggestive of increased B7-H3 expression being related to a Th17 type response (as seen in animal models) and possibly with an increase in regulatory T cells.4 As B7-H3 directed therapy enters into a clinical study in human prostate cancer it will be critical to perform correlative science to help elucidate which molecular interactions can be capitalized on, particularly for combination therapy trials.
In summary, we find that B7-H3 expression is present in the majority of localized prostate cancer cases, with levels of B7-H3 increasing with tumor extent and aggressiveness. B7-H3 is potentially androgen regulated and related to the TGF-β signaling pathway, however, these associations require further study. B7-H3 expression characteristics may help guide patient selection for clinical trials, particularly to anti-B7-H3 antibodies such as enoblituzumab.
Written by: B Benzon1, SG Zhao2, MC Haffner1,3, M Takhar4, N Erho4, K Yousefi4, P Hurley1, JL Bishop5, J Tosoian1, K Ghabili1, M Alshalalfa4, S Glavaris1, BW Simons1, P Tran1,6, E Davicioni4, RJ Karnes5,7, K Boudadi1,8, ES Antonarakis1,8, EM Schaeffer9, CG Drake1,8, F Feng2 and AE Ross1
1 Thompson RH, Kwon ED, Allison JP. Inhibitors of B7-CD28 costimulation in urologic malignancies. Immunotherapy 2009; 1: 129–139.
3 Chapoval AI, Ni J, Lau JS, Wilcox RA, Flies DB, Liu D et al. B7-H3: a costimulatory molecule for T cell activation and IFN-gamma production. Nat Immunol 2001; 2: 269–274.
4 Luo L, Zhu G, Xu H, Yao S, Zhou G, Zhu Y et al. B7-H3 Promotes pathogenesis of autoimmune disease and inflammation by regulating the activity of different T cell subsets. PLoS ONE 2015; 10: e0130126.
5 Kreymborg K, Haak S, Murali R, Wei J, Waitz R, Gasteiger G et al. Ablation of B7-H3 but not B7-H4 results in highly increased tumor burden in a murine model of spontaneous prostate cancer. Cancer Immunol Res 2015; 3: 849–854.
6 Sun X, Vale M, Leung E, Kanwar JR, Gupta R, Krissansen GW. Mouse B7-H3 induces antitumor immunity. Gene Ther 2003; 10: 1728–1734.
7 Luo L, Chapoval AI, Flies DB, Zhu G, Hirano F, Wang S et al. B7-H3 enhances tumor immunity in vivo by costimulating rapid clonal expansion of antigen-specific CD8+ cytolytic T cells. J Immunol 2004; 173: 5445–5450.
8 Roth TJ, Sheinin Y, Lohse CM, Kuntz SM, Frigola X, Inman BA et al. B7-H3 ligand expression by prostate cancer: a novel marker of prognosis and potential target for therapy. Cancer Res 2007; 67: 7893–7900.
9 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 USA 2007; 104: 19458–19463.
10 Parker AS, Heckman MG, Sheinin Y, Wu KJ, Hilton TW, Diehl NN et al. Evaluation of B7-H3 expression as a biomarker of biochemical recurrence after salvage radiation therapy for recurrent prostate cancer. Int J Radiat Oncol Biol Phys 2011; 79: 1343–1349.
11 Chavin G, Sheinin Y, Crispen PL, Boorjian SA, Roth TJ, Rangel L et al. Expression of immunosuppresive B7-H3 ligand by hormone-treated prostate cancer tumors and metastases. Clin Cancer Res 2009; 15: 2174–2180.
12 Yamoah K, Johnson MH, Choeurng V, Faisal FA, Yousefi K, Haddad Z et al. Novel biomarker signature that may predict aggressive disease in african american men with prostate cancer. J Clin Oncol 2015; 33: 2789–2796.
13 Ross AE, Johnson MH, Yousefi K, Davicioni E, Netto GJ, Marchionni L et al. Tissue-based genomics augments post-prostatectomy risk stratification in a natural history cohort of intermediate- and high-risk men. Eur Urol 2016; 69: 157–165.
14 Erho N, Crisan A, Vergara IA, Mitra AP, Ghadessi M, Buerki C et al. Discovery and validation of a prostate cancer genomic classifier that predicts early metastasis following radical prostatectomy. PLoS ONE 2013; 8: e66855.
15 Piccolo SR, Sun Y, Campbell JD, Lenburg ME, Bild AH, Johnson WE. A singlesample microarray normalization method to facilitate personalized-medicine workflows. Genomics 2012; 100: 337–344.
16 Yu J, Yu J, Mani RS, Cao Q, Brenner CJ, Cao X et al. An integrated network of androgen receptor, polycomb, and TMPRSS2-ERG gene fusions in prostate cancer progression. Cancer Cell 2010; 17: 443–454.
17 Robinson JT, Thorvaldsdottir H, Winckler W, Guttman M, Lander ES, Getz G et al. Integrative genomics viewer. Nat Biotechnol 2011; 29: 24–26.
18 Haffner MC, Aryee MJ, Toubaji A, Esopi DM, Albadine R, Gurel B et al. Androgeninduced TOP2B-mediated double-strand breaks and prostate cancer gene rearrangements. Nat Genet 2010; 42: 668–675.
19 Therneau TM, Li H. Computing the Cox model for case cohort designs. Lifetime Data Anal 1999; 5: 99–112.
20 Tomlins SA, Alshalalfa M, Davicioni E, Erho N, Yousefi K, Zhao S et al. Characterization of 1577 primary prostate cancers reveals novel biological and clinicopathologic insights into molecular subtypes. Eur Urol 2015; 68: 555–567.
21 Dennis G Jr, Sherman BT, Hosack DA, Yang J, Gao W, Lane HC et al. DAVID: Database for Annotation, Visualization, and Integrated Discovery. Genome Biol 2003; 4: P3.
22 Agell L, Hernandez S, Nonell L, Lorenzo M, Puigdecanet E, de Muga S et al. A 12-gene expression signature is associated with aggressive histological in prostate cancer: SEC14L1 and TCEB1 genes are potential markers of progression. Am J Pathol 2012; 181: 1585–1594.
24 Powles T, Eder JP, Fine GD, Braiteh FS, Loriot Y, Cruz C et al. MPDL3280A (anti-PD-L1) treatment leads to clinical activity in metastatic bladder cancer. Nature 2014; 515: 558–562.
25 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 JMed 2015; 372: 2018–2028.
26 Kantoff PW, Higano CS, Shore ND, Berger ER, Small EJ, Penson DF et al. Sipuleucel-T immunotherapy for castration-resistant prostate cancer. N Engl JMed 2010; 363: 411–422.
27 Kwon ED, Drake CG, Scher HI, Fizazi K, Bossi A, van den Eertwegh AJ et al. Ipilimumab versus placebo after radiotherapy in patients with metastatic castration-resistant prostate cancer that had progressed after docetaxel chemotherapy (CA184-043): a multicentre, randomised, double-blind, phase 3 trial. Lancet Oncol 2014; 15: 700–712.
28 Powderly J, Cole G, Flaherty K, Szmulewitz RZ, Ribas A, Weber J et al. Interim results of an ongoing Phase I, dose escalation study of MGA271 (Fc-optimized humanized anti-B7-H3 monoclonal antiody) in patients with refractory B7-H3- expressing neoplamss whose vasculature expresses B7-H3. Journal for Immunotherapy of Cancer. 2015; 3(Suppl 2).
29 Lupu CM, Eisenbach C, Lupu AD, Kuefner MA, Hoyler B, Stremmel W et al. Adenoviral B7-H3 therapy induces tumor specific immune responses and reduces secondary metastasis in a murine model of colon cancer. Oncol Rep 2007; 18: 745–748.
31 Wu CP, Jiang JT, Tan M, Zhu YB, Ji M, Xu KF et al. Relationship between co-stimulatory molecule B7-H3 expression and gastric carcinoma histology and prognosis. World J Gastroenterol 2006; 12: 457–459.
32 Loos M, Hedderich DM, Ottenhausen M, Giese NA, Laschinger M, Esposito I et al. Expression of the costimulatory molecule B7-H3 is associated with prolonged survival in human pancreatic cancer. BMC Cancer 2009; 9: 463.
33 Crispen PL, Sheinin Y, Roth TJ, Lohse CM, Kuntz SM, Frigola X et al. Tumor cell and tumor vasculature expression of B7-H3 predict survival in clear cell renal cell carcinoma. Clin Cancer Res 2008; 14: 5150–5157.
34 Chaudhry A, Rudensky AY. Control of inflammation by integration of environmental cues by regulatory T cells. J Clin Invest 2013; 123: 939–944.
35 Galon J, Costes A, Sanchez-Cabo F, Kirilovsky A, Mlecnik B, Lagorce-Pages C et al. Type, density, and location of immune cells within human colorectal tumors predict clinical outcome. Science 2006; 313: 1960–1964.