Abstract
Patients with prostate cancer (PC) generally do not respond favorably to immune checkpoint inhibitors, which may be due to a low abundance of tumor-infiltrating lymphocytes even when mutational load is high. Here, we identified a patient who presented with high-grade primary prostate cancer with two adjacent tumor nodules. While both nodules were mismatch repair-deficient (MMRd), exhibited pathogenic MSH2 and MSH6 alterations, had a high tumor mutational burden (TMB), and demonstrated high microsatellite instability (MSI), they had markedly distinct immune phenotypes. The first displayed a dense infiltrate of lymphocytes (“hot nodule”), while the second displayed significantly fewer infiltrating lymphocytes (“cold nodule”). Whole-exome DNA analysis found that both nodules shared many identical mutations, indicating that they were derived from a single clone. However, the cold nodule appeared to be sub-clonal relative to the hot nodule, suggesting divergent evolution of the cold nodule from the hot nodule. Whole-transcriptome RNA analysis found that the cold nodule demonstrated lower expression of genes related to antigen presentation (HLA) and, paradoxically, classical tumor immune tolerance markers such as PD-L1 (CD274) and CTLA-4. Immune cell deconvolution suggested that the hot nodule was enriched not only in CD8+ and CD4 + T lymphocytes, but also in M1 macrophages, activated NK cells, and γδ T cells compared to the cold nodule. This case highlights that MMRd/TMB-high PC can evolve to minimize an anti-tumor immune response, and nominates downregulation of antigen presentation machinery (HLA loss) as a potential mechanism of adaptive immune evasion in PC.
Similar content being viewed by others
Introduction
Prostate cancer (PC) is the most prevalent cancer in men and remains the second most common cause of cancer-related mortality1. While most deaths occur due to metastatic disease, treatment for localized PC is curative in many patients and the 5-year survival rate is greater than 95%2. In both early and advanced settings of PC, the androgen receptor (AR) remains a critical driver and therapeutic target3,4,5. Androgen deprivation therapies (ADT) and AR-targeted therapies (ART) are two therapeutic strategies that disrupt oncogenic signaling associated with the AR. While these therapies extend survival in patients with PC, resistance to ADTs and ARTs is inevitable. Several alternative targeted therapies are also being developed including those that target DNA damage repair (PARP inhibitors)6,7,8, PTEN loss (AKT inhibitors)9, and radiotherapies that target PSMA (Lu177-PSMA-617)10. In addition to these molecularly-targeted therapies, other strategies are leveraging a patient’s own immune system to treat their cancer. However, despite favorable outcomes in other cancer types, the use of immune checkpoint inhibitors (ICIs) and other immunotherapeutic strategies (including cell-based therapies and cancer vaccines) have proven less effective in PC11.
In broad oncology practice, clinical predictive biomarkers for ICI sensitivity include PD-L1 expression, TMB greater than 10 mutations per megabase (10 muts/Mb) and MSI-high status12,13. However, these markers are observed in only 3–5% of patients with PC14. The majority of PC tumors are regarded as immunologically cold, meaning that they display a low abundance of tumor infiltrating T-cells. This is thought to be one reason that patients with PC have poor response towards current immune-directed therapeutic strategies15,16. Due to the very small population of patients with PC who successfully mount anti-tumor immunity, it has been a challenge to identify key features that enable or suppress immune responses toward PC.
Here, we have identified a unique case of a patient with high-risk localized PC with two adjacent tumor nodules displaying distinctly divergent infiltrating immune-cell phenotypes. Despite both nodules being TMB-high and MSI-high, one nodule demonstrated a markedly higher density of tumor-infiltrating T cells. We interrogated these two tumor nodules to discern molecular differences between their exomes, transcriptomes, and tumor microenvironments to identify possible mechanisms that regulate immune-cell interactions and immune evasion in PC.
Results
Case report
A 61-year-old man was found to have a prostate-specific antigen (PSA) level of 10.4 ng/mL and a palpable prostate nodule that was clinical stage cT2b. A prostatic biopsy revealed adenocarcinoma in 5 of 12 cores, with Gleason grade 5 + 5 = 10. Computed tomography of the chest, abdomen, and pelvis and bone scintigraphy were negative for local or distant metastases. Magnetic resonance imaging of the prostate gland showed the entire left peripheral zone occupied by tumor with central necrosis, which deformed the capsule and abutted the left neurovascular bundle without frank invasion (Fig. 1a). Given the diagnosis of high-risk localized PC, the patient elected to enroll in a clinical trial of neoadjuvant treatment prior to radical prostatectomy (RP). The trial randomized patients with high-risk PC to treatment with androgen-deprivation therapy (ADT; degarelix 240 mg subcutaneously) given two weeks prior to RP (Cohort A) or to cyclophosphamide (200 mg/m2 intravenously) and a Granulocyte-macrophage colony-stimulating factor (GM-CSF)-secreting allogeneic cellular vaccine (GVAX; composed of 2.5 × 108 PC3 cells and 1.6 × 108 LNCaP cells, injected intradermally) plus ADT given two weeks prior to RP (Cohort B). The main objective of that trial was to analyze CD8 + T cell density in the tumor obtained at RP (NCT01696877)17. This patient was randomized to cohort B and was treated with cyclophosphamide/GVAX plus ADT followed by RP. His PSA level became undetectable post-operatively and remains undetectable at the time of writing, greater than 7 years after prostatectomy. This patient has achieved a 7-year cancer-free interval thus far despite having both hot and cold nodules present in his prostate. This suggests that even the cold nodule exhibited a favorable response to combination therapy consisting of RP, ADT and GVAX vaccination, likely due to the fact that this nodule also showed MMR deficiency and hypermutation.
On pathologic examination of the patient’s RP specimen (Supplementary Fig. 1), there was invasive high-grade prostate adenocarcinoma that occupied relatively large areas of the peripheral zone of the prostate gland. Upon medium-power view, two nearby yet distinct appearing nodules were present in the left peripheral zone that were separated by an intervening stromal component (Supplementary Fig. 1A). Each nodule displayed high-grade prostate adenocarcinoma (Gleason 4 + 5 = 9), but one showed a higher density of infiltrating mononuclear inflammatory cells consistent with lymphocytes (termed the “hot” nodule) compared with the other nodule (termed the “cold” nodule) (Fig. 1b, c). By single-plex chromogenic immunohistochemistry, the hot nodule showed a much greater density of CD3+ and CD8+ tumor-infiltrating T cells (TILs) than the cold nodule (Fig. 1b, c and Supplementary Fig. 1D–I). The increased number of CD3 and CD8 cells staining in the hot nodule was supported by our quantification in which we determined the density of positive cells per mm2 (Fig. 1d). In addition to there being a relatively high density of CD8+ and CD3 + T cells in the tumor interior, the hot nodule also contained dense collections of these cells in a band-like pattern toward the periphery of the tumor (Fig. 1b and Supplementary Fig. 1D)—a highly unusual feature for PC. Given that TIL density is an established biomarker of immunologic recognition and anti-tumor18, we sought to characterize molecular differences between these nodules that could inform mechanisms of immune stimulation in the hot nodule or immune evasion in the cold nodule.
Genomic analysis
With such striking differences observed in immune-cell densities on histology, we anticipated divergent (or separate) genomic profiles between these nodules. However, genomic sequencing of bulk tumor tissue using Personal Genome Diagnostics (PGDx) showed that both the hot and cold nodules demonstrated evidence of MMRd and extremely high TMB. As compared with aggregated data from PC which has a median TMB of 2–3 muts/Mb, in the hot nodule the TMB was 131 muts/Mb, the frameshift burden was 12 frameshifts/Mb, and the frameshift proportion was high at 9%. Similarly, in the cold nodule the TMB was 164 muts/Mb, the frameshift burden was 15 frameshifts/Mb, and the frameshift proportion was 9%. In a comparative analysis relative to all other cancer types in The Cancer Genome Atlas (TCGA) dataset, these nodules exhibited the second and third highest TMB relative to all other primary PC tumors (n = 495) and the second and sixth highest TMB relative to all other patients (n = 468) even amongst mutation-high cancers such as melanomas (Supplementary Figure 2). Furthermore, both nodules harbored inactivating mutations in MSH2 and MSH6 (MSH2 p.E809*, MSH6 p.F1104Lfs*11) with loss of protein expression by immunohistochemistry, confirming MMRd status, and were MSI-high. Germline genetic testing was unremarkable, confirming that these mutations were acquired somatic events. Both nodules also had POLE mutations (p.Y2003C, p.A782V, and p.D756G) although none were in the exonuclease domain. Additional mutations identified and annotated with Ensembl’s variant effect predictor (VEP)19 are listed in Supplementary Table 1. Due to the unexpected finding of MMRd and high-TMB in both tumor nodules, we sought to further comprehensively characterize additional molecular features that may explain the distinct TMEs seen in the two nodules.
To this end, we first examined COSMIC mutational signatures in the two tumors including insertion and deletions (InDel), single-base substitutions, and double-base substitutions between the hot and cold nodules (Fig. 1e)20. In this regard, the two nodules showed a high degree of similarity between their mutational processes at the nucleotide level, providing strong evidence that they were clonally related. We then examined the relative contribution of each signature type in the two nodules (Fig. 1f); again, both nodules showed predominantly MMRd-related mutational patterns. Interestingly, the hot nodule displayed a small relative contribution of defective homologous recombination repair (HRR) that was not observed in the cold nodule, even though comparable variant allele frequencies (VAFs) of BRCA1/2 alterations were observed and none of them were pathogenic. We next sought to examine the landscape of VAFs in both nodules; here, we found unique patterns in the hot nodule. Specifically, the hot nodule exclusively showed unique pathogenic mutations in TP53 and PTEN (Fig. 1g), although both of these mutations were of low variant allele frequency. Further, the hot nodule was composed of a single tumor clone while the cold nodule was composed of two clones (Fig. 1h). We next modeled the order in which mutations accumulated through a pseudo-time analysis using the VAFs of all detectable gene alterations. Based on this approach, the cold nodule initially accumulated mutations quicker and at an earlier time point compared to the hot nodule (Fig. 1l). At a later pseudo-time, the cold nodule exhibited a similar overall rate of variant accumulation as the hot but never acquired the unique pathogenic TP53 and PTEN mutations. While the overarching mutational processes appeared very similar, these nuanced genomics analyses indicated that these tumors exhibited distinct rates of VAF accumulation and harbored certain unique mutations.
Transcriptomic analysis
We next sought to examine the whole transcriptomes (Fig. 2) of the hot and cold nodules, to interrogate potential differences in specific genes or overarching biological processes. We performed gene set enrichment analysis (GSEA) of the 50 hallmark signatures21,22, and expectedly found that the hallmark signatures enriched in the hot nodule were primarily related to immune-regulatory signaling pathways such as IFN-γ/α and TNF-α (Fig. 2a, b, c). Gene expression and genomic alteration status of each gene in the signatures are included as supplementary data (Supplementary Tables 4, 5). Otherwise, based on a Pearson correlation of relative transcript abundance, these tumors displayed strikingly similar overall transcriptomes (R2 = 0.928, p value < 0.0001) (Fig. 2d). Yet the hot nodule was enriched for a variety of immunoglobulin genes, including IGLV3-1 and IGHV3-23 among others (Fig. 2d, f). Interestingly, the hot nodule was also enriched in signatures associated with therapy resistance such as epithelial-to-mesenchymal transition (EMT)23 and WNT signaling24. While androgen signaling is thought to negatively impact the activity of immune cells25,26, the hallmark androgen response signature was not enriched in either the hot or cold nodule, suggesting that both had similar expression of AR-related genes (Fig. 2c, d, h). Despite this limited overall difference in the signaling pathways, individual AR-related and NEPC genes were all generally upregulated in the cold tumor (Fig. 2h, l).
When examining the expression of specific individual genes, the hot nodule displayed increased expression of many HLA27 genes including HLA-B, B2M, HLA-DRA and CD74 compared to the cold nodule (Fig. 2e). When we examined epigenetic regulators that regulate HLA expression28, the hot nodule contained a unique DNMT3B variant while the cold nodule had a unique DNMT3AP1 variant. Based on gene expression patterns, TAP1, TAP2, and B2M were all decreased in the cold nodule compared to the hot nodule. Similarly, the hot nodule was enriched in many immunoglobulin genes, which may be due to differences in relative abundance of B-cells and plasma cells. However, differences between the nodules are not robust when examined by CIBERSORT or IHC and therefore this does not directly explain the enrichment of immunoglobulin genes in the hot nodule (Fig. 2f, Supplementary Fig. 3). The hot nodule also had increased levels of ICI target genes such as PD-1, PD-L1 and CTLA4. (Fig. 2g). While multiple immune-related genes had seemingly distinct expression patterns, there were limited differences when we examined critical regulators of AR signaling such as FOXA1, HOXB1329,30, and KLK2/331,32 (Fig. 2h) or neuroendocrine PC regulatory genes such as INSM1 and ASCL1 (Fig. 2i). Altogether, these nuanced transcriptional analyses uncovered key transcriptional differences (characterized by loss of HLA gene expression and reduced immunoglobulin gene expression in the cold nodule) despite a seemingly similar broad transcriptional profile in the two nodules.
Immune cell deconvolution
We next sought to comprehensively interrogate the TME in the two nodules, through inferential analysis via bulk RNA sequencing data. When evaluating projected abundance and relative fraction of immune cells using CIBERSORT33,34, the hot nodule exhibited higher levels of CD8+ and CD4 + T lymphocytes, as well as more M1 macrophages, activated natural killer (NK) cells, and gamma delta (γδ) T cells compared to the cold nodule (Fig. 3a, b). Other immune cell subtypes from CIBERSORT are summarized in Supplementary Tables 2 and 3. We next sought to explore if the hot nodule had differential immune-cell fractions when compared to other patients in the original GVAX clinical trial17. Using Nanostring IO360 gene expression data from a selective panel of immune-related genes (previously generated17), the absolute abundance of immune-cell subtypes was inferred using CIBERSORT33,34 and compared between patients enrolled on the trial (n = 43). To this end, the hot nodule from our patient displayed higher levels of many immune-cell subtypes relative to all other patients in the trial regardless of the treatment arm (Fig. 3c). Of note, Nanostring IO360 gene expression data from the cold nodule were not previously generated and could not be reported. Taken together, these immune-cell deconvolution analyses indicated that the hot nodule displayed a unique TME compared to all other PC patients in the GVAX clinical trial across all three arms.
Discussion
TMB-high and MSI-high status in tumors are associated with increased immunogenicity, immune recognition, and elimination of tumor antigens by the immune system. Due to these factors, TMB and MSI status are currently used as clinical biomarkers to deploy ICI treatments in a histology-agnostic fashion35,36. Our study demonstrates that additional features may regulate the TME, and further investigations are required to identify mechanisms that drive immune-cell interaction in the 3–5% of PC tumors that are TMB-high and/or MSI-high. As indicated by two adjacent tumor nodules in this PC patient, the hot nodule had higher expression of HLA and immunoglobulin genes compared to the cold nodule. Or perhaps, viewed alternatively, the cold nodule lost HLA and immunoglobulin gene expression as a means of adaptive immune resistance. An alternative hypothesis is that the lower expression of HLA and immunoglobulin genes could have been the result of a sparser immune-cell infiltrate in the cold nodule due to the bulk RNA sequencing methodology used (rather than single-cell RNA sequencing). Interestingly, some recent studies have indicated that tumors can decrease expression of antigen-presentation genes in order to evade immune-cell surveillance27,37. As another possibility, each nodule also harbored alterations in regulatory genes and epigenetic factors that regulate immune surveillance28,38. This may have contributed to the reduced immune surveillance functions in the cold nodule. We also found that both nodules exhibited largely similar mutational processes, with the exception of a component of HRR, but had distinct numbers of subclones and VAF accumulation rates.
Of note, our GSEA analysis indicated that the hot nodule was interestingly enriched in EMT and WNT signaling, two pathways typically associated with the immune evasion39,40. Our study implicates other pathways may have led to the distinct immune cell phenotypes in these nodules. With respect to ICI implications, our findings indicate that clinical-grade molecular tests should consider reporting immune-regulatory signatures and relative expression levels of HLA and immunoglobulin genes in addition to TMB and MSI status.
Currently, the low response rates of heterogeneous solid tumors, including PC41, to single-agent immunotherapies has led to active investigation of combination strategies. Our patient was enrolled in a trial in which ADT was combined with a cellular PC vaccine, GVAX17. Interestingly, both the hot and cold nodules shared variants with GVAX based on the variant profiles of PC3 and LNCaP cells found in the DepMap database. While the hot nodule overall shared less variants with these cell lines (Supplementary Table 6), it remains possible that specific highly-immunogenic variant(s) produced antigens that mediated the inflamed response. Previous work has demonstrated that androgens may suppress T-cell function26, and that inhibition of AR (via ADT and/or ART) may alter the TME by enhancing T-cell function25 The rationale of the GVAX trial was that ADT would complement the vaccine to promote infiltration of pro-inflammatory immune cells and ultimately CD8 + T-cells into tumors42. In this patient, both nodules underwent the same systemic therapy, which indicates that mechanisms other than ADT drove the significant differences in the TMEs. Stromal cells may also contribute to the divergent immune phenotypes43. The cold nodule was visually associated with greater proportions of stromal cells. However, functional mapping with single cell resolution (which was not possible here) is required to determine if specific stromal cells may have contributed to the differences in immune cell infiltration. A more refined understanding, which will require more patient studies, is necessary to identify the recurrent mechanisms that alter the TME in PC patients, which may impact the deployment of immune checkpoint blockade, cell-based therapies, and vaccine therapies in solid tumors.
We also demonstrate here that a prostate tumor may interact with many forms of immune cells in its microenvironment. Outside of regulatory and helper T-cells, we inferred enrichment of γδ T-cells, M1 macrophages, and activated NK cells in the hot nodule relative to the cold nodule. Studies have shown that NK44 or γδ T-cells45,46 are able to invade and kill tumors. As compared to T-cells, which have been less effective in driving tumor response in solid tumors, further research should measure the relative infiltration rates of NK and/or γδ T cells in the tumor microenvironment, which may allow us to consider such cell-based therapies for treating prostate tumors47.
There are technical and statistical limitations based on this case study of two nodules. We have made descriptive observations based on thresholds that would typically be considered notable in clinical reports or larger patient cohorts. However, validating function and causality would still require further modeling of the patients’ tumors. To generalize our key observations, we would also require validation of key findings in greater cohorts of dMMR tumors.
In conclusion, our case study highlights that complex genomic and transcriptomic alterations can regulate the TME in TMB-high and MSI-high PC tumors. Therefore, in addition to examining mutational burdens and signatures, it is of relevance to interrogate additional immune-regulatory and clonal processes, as well as to consider the types of immune cells that are associated with the tumor and its microenvironment. Finally, the observation of HLA loss in the context of MMR-deficient cancers observed here (and reported elsewhere)48 may have broader relevance as an immune evasion mechanism in other tumor contexts. In this patient with an ultra-high TMB in both the hot and cold nodules, it is tempting to speculate that the immunogenic effect of accumulating mutations may have been counterbalanced by transcriptional events that dampened immunoreactivity.
Methods
Patient approval
The patient provided written consent for inclusion in the clinical trial, for further genomic analysis of their tumor as presented herein, and to publish the findings.
Ethics approval
This study has complied with all relevant ethical regulations including the Declaration of Helsinki. IRB approval was obtained for the data related to the initial clinical trial and further details can be found in the original manuscript publication17. Patient consent was acquired for the additional sequencing on the two tumor nodules and the analyses included in this manuscript. All other data included in this study was derived from public resources and these resources are provided in the manuscript. Ethical approval from the source datasets included can be found in the respective sources.
Sample preparation and sequencing
Sample preparation, library construction, exome capture, and next generation sequencing of FFPE tumor and normal saliva samples were performed at Personal Genome Diagnostics (Baltimore, MD). Total DNA and RNA samples were isolated from patient FFPE tumor tissue specimens and sequencing was performed using the HiSeq2500 system with 100 base pair reads at Personal Genome Diagnostics (Baltimore, MD). Further details of sample preparation procedures and sequencing parameters can be found in a prior study by Jones, S. et al.49.
Whole Exome Sequencing (WES) Data seq processing
Raw fastq files were obtained and mapped to the GRCh38 reference genome using BWA MEM50. Picard tools were used for BAM sorting and marking duplicate reads. Base quality score recalibration was done using GATK51. Mutect2 was used for somatic variant calling, with a patient matched normal saliva sample used as input alongside the tumor samples. Tumor VCF files were annotated with Ensembl VEP19. These WES data preprocessing steps were conducted using Minnesota Supercomputing Institute High Performance Computing resources on a CentOS Linux environment. Mutational signatures were analyzed using the MutationalPatterns20 package in R, with which COSMIC mutational signatures were used to infer DNA damage signatures52. The maftools package in R was used to generate and plot annotated VAF results. The tcgaCompare function from the maftools R package was used to compute the tumor mutational burden (TMB) of the two prostate nodule samples and compare them to 33 different TCGA datasets. Details regarding the TCGA studies included in the function and in our analysis can be found here: https://github.com/PoisonAlien/TCGAmutations.
Microsatellite instability (MSI) status and tumor mutational burden (TMB) were defined as previously described53. Frameshift mutation burden was defined as the number of insertion/deletion frameshift mutations per Megabase of DNA (muts/Mb), as previously described54. Frameshift mutation proportion was defined as the proportion of frameshift mutations relative to all nonsynonymous sequence alterations per Megabase of DNA (muts/Mb), as previously described54.
Whole transcriptome sequencing (WTS) data processing
Transcriptome sequence data processing and analysis were performed using pipelines at the Minnesota Supercomputing Institute and University of Minnesota Informatics Institute (UMII) at the University of Minnesota. Raw reads were trimmed, aligned to the GRCh38 human genome, and gene-level read counts were generated using the CHURP pipeline55. All downstream gene expression analyses and visualizations were conducted using R (4.2.1), RStudio (2022.07.2 + 576) and GraphPad Prism 9. Transcripts per million (TPM) were calculated from raw counts for use in all downstream transcriptomic analyses and visualizations.
Nanostring data processing
Normalized count gene expression data from the Nanostring IO360 immune gene panel (https://nanostring.com/products/ncounter-assays-panels/oncology/pancancer-immune-profiling/) consisting of 730 genes was obtained for 43 patients from the GVAX trial, including the hot tumor nodule from the patient of interest (sample 15-21233).
Gene set enrichment analysis (GSEA)
The differences in transcripts per million for each gene was calculated and divided by the mean expression values in T1 and T2. These values were used to generate a ranked profile in order to conduct the subsequent pre-ranked GSEA using Hallmark Signatures21.
CIBERSORT analysis
CIBERSORTx33,34 was used to estimate immune cell fractions and abundances. The LM22 microarray dataset, consisting of 547 genes that distinguish 22 mature human hematopoietic populations, was used as the signature matrix input for the imputation of immune cell fractions in the tumor samples. Normalized whole-transcriptome data were used to compare immune cell fractions and abundance of the two prostate nodules. Nanostring data were used to compare immune cell abundance across 43 patients in the GVAX trial.
Statistics and reproducibility
Pairwise gene correlation coefficients (Pearson’s), p values, and adjusted p value statistics were calculated using R 4.2.2 (R Core Team; 2022), the stats (v4.2.2; R Core Team; 2022), the Tidyverse (v1.3.2; Wickham; 2019), and the Hmsic (v4.7-2; Harrell Jr F; 2022) packages.
IHC and image analysis for CD8 Cell IHC density measurements
IHC staining for CD8 was performed previously as a single-plex assay with DAB as chromogen (PMID: 32173650). The whole slide image was uploaded to HALO 3.6.4134.137 (Indica Labs). Each region (hot nodule and cold nodule) was separately annotated. The HALO random forest algorithm was used to train a classifier to detect total tissue area in the annotated region, which excluded lumens and bare glass areas. The Cytonuclear 2.0.9 algorithm was used to segment the cells and analyze the CD8 positively stained cells. The cell segmentation was based on hematoxylin staining and the following parameters: nuclear contrast threshold, minimal nuclear optical density, nuclear size and nuclear segmentation aggressiveness.
Reporting summary
Further information on research design is available in the Nature Research Reporting Summary linked to this article.
Data availability
Both the hot and the cold nodules tumor from this patient underwent bulk whole-exome DNA sequencing (WES) and bulk whole-transcriptome sequencing (WTS) at PGDx (Baltimore, MD). The data corresponding to this study is now available at the SRA (BioProject Accession number PRJNA1032970) and GEO (Accession number GSE246684). The additional Nanostring data from the original clinical trial that was analyzed during this study is included in a published article and its supplementary files (https://doi.org/10.1158/1078-0432.CCR-19-3372)17.
Code availability
The basic operational source code has been deposited into the following Zenodo Digital Repository (https://doi.org/10.5281/zenodo.10206909). Additional requests regarding codes and scripts can be addressed to the corresponding author.
References
Siegel, R. L., Miller, K. D., Wagle, N. S. & Jemal, A. Cancer statistics, 2023. CA Cancer J. Clin. 73, 17–48 (2023).
SEER*Explorer: An interactive website for SEER cancer statistics [Internet]. Surveillance Research Program, National Cancer Institute; 2023 Apr 19. [updated: 2023 Nov 16; cited 2024 Jan 19]. Available from: https://seer.cancer.gov/statistics-network/explorer/. Data source(s): SEER Incidence Data, November 2022 Submission (1975-2020), SEER 22 registries (excluding Illinois and Massachusetts). Expected Survival Life Tables by Socio-Economic Standards.
Antonarakis, E. S. et al. AR-V7 and resistance to enzalutamide and abiraterone in prostate cancer. N. Engl. J. Med. 371, 1028–1038 (2014).
Abida, W. et al. Genomic correlates of clinical outcome in advanced prostate cancer. Proc. Natl. Acad. Sci. USA 116, 11428–11436 (2019).
Quigley, D. A. et al. Genomic Hallmarks and structural variation in metastatic prostate cancer. Cell 174, 758–769.e759 (2018).
Nizialek, E. & Antonarakis, E. S. PARP inhibitors in metastatic prostate cancer: evidence to date. Cancer Manag. Res. 12, 8105–8114 (2020).
de Bono, J. et al. Olaparib for metastatic castration-resistant prostate cancer. N. Engl. J. Med. 382, 2091–2102 (2020).
Mateo, J. et al. A decade of clinical development of PARP inhibitors in perspective. Ann. Oncol. 30, 1437–1447 (2019).
Toren, P. & Zoubeidi, A. Targeting the PI3K/Akt pathway in prostate cancer: challenges and opportunities (review). Int J. Oncol. 45, 1793–1801 (2014).
Sartor, O. et al. Lutetium-177-PSMA-617 for metastatic castration-resistant prostate cancer. N. Engl. J. Med. 385, 1091–1103 (2021).
Antonarakis, E. S. et al. Pembrolizumab for treatment-refractory metastatic castration-resistant prostate cancer: multicohort, open-label Phase II KEYNOTE-199 study. J. Clin. Oncol. 38, 395–405 (2020).
Chan, T. A. et al. Development of tumor mutation burden as an immunotherapy biomarker: utility for the oncology clinic. Ann. Oncol. 30, 44–56 (2019).
Haffner, M. C. et al. Comprehensive evaluation of programmed Death-Ligand 1 expression in primary and metastatic prostate cancer. Am. J. Pathol. 188, 1478–1485 (2018).
Abida, W. et al. Analysis of the prevalence of microsatellite instability in prostate cancer and response to immune checkpoint blockade. JAMA Oncol. 5, 471–478 (2019).
Cha, H. R., Lee, J. H. & Ponnazhagan, S. Revisiting immunotherapy: a focus on prostate cancer. Cancer Res. 80, 1615–1623 (2020).
Sena, L. A., Denmeade, S. R. & Antonarakis, E. S. Targeting the spectrum of immune checkpoints in prostate cancer. Expert Rev. Clin. Pharm. 14, 1253–1266 (2021).
Obradovic, A. Z. et al. T-Cell infiltration and adaptive treg resistance in response to androgen deprivation with or without vaccination in localized prostate cancer. Clin. Cancer Res. 26, 3182–3192 (2020).
Tumeh, P. C. et al. PD-1 blockade induces responses by inhibiting adaptive immune resistance. Nature 515, 568–571 (2014).
McLaren, W. et al. The ensembl variant effect predictor. Genome Biol. 17, 122 (2016).
Manders, F. et al. MutationalPatterns: the one stop shop for the analysis of mutational processes. BMC Genom. 23, 134 (2022).
Subramanian, A. et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl. Acad. Sci. USA 102, 15545–15550 (2005).
Liberzon, A. et al. The Molecular Signatures Database (MSigDB) hallmark gene set collection. Cell Syst. 1, 417–425 (2015).
Cancer of Any Site - Cancer Stat Facts. in Surveillance, Epidemiology, and End Results (SEER) Program Populations (1969-2019), 2021 (National Cancer Institute, 2021).
Murillo-Garzon, V. & Kypta, R. WNT signalling in prostate cancer. Nat. Rev. Urol. 14, 683–696 (2017).
Guan, X. et al. Androgen receptor activity in T cells limits checkpoint blockade efficacy. Nature 606, 791–796 (2022).
Kissick, H. T. et al. Androgens alter T-cell immunity by inhibiting T-helper 1 differentiation. Proc. Natl. Acad. Sci. USA 111, 9887–9892 (2014).
Sadagopan, A., Michelakos, T., Boyiadzis, G., Ferrone, C. & Ferrone, S. Human leukocyte antigen Class I Antigen-processing machinery upregulation by anticancer therapies in the era of checkpoint inhibitors: a review. JAMA Oncol. 8, 462–473 (2022).
Rodems, T. S. et al. Reversible epigenetic alterations regulate class I HLA loss in prostate cancer. Commun. Biol. 5, 897 (2022).
Pomerantz, M. M. et al. The androgen receptor cistrome is extensively reprogrammed in human prostate tumorigenesis. Nat. Genet. 47, 1346–1351 (2015).
Pomerantz, M. M. et al. Prostate cancer reactivates developmental epigenomic programs during metastatic progression. Nat. Genet. 52, 790–799 (2020).
Chen, S. et al. Single-cell analysis reveals transcriptomic remodellings in distinct cell types that contribute to human prostate cancer progression. Nat. Cell Biol. 23, 87–98 (2021).
Sun, Z., Pan, J. & Balk, S. P. Androgen receptor-associated protein complex binds upstream of the androgen-responsive elements in the promoters of human prostate-specific antigen and kallikrein 2 genes. Nucleic Acids Res. 25, 3318–3325 (1997).
Newman, A. M. et al. Determining cell type abundance and expression from bulk tissues with digital cytometry. Nat. Biotechnol. 37, 773–782 (2019).
Newman, A. M. et al. Robust enumeration of cell subsets from tissue expression profiles. Nat. Methods 12, 453–457 (2015).
Jardim, D. L., Goodman, A., de Melo Gagliato, D. & Kurzrock, R. The challenges of tumor mutational burden as an immunotherapy biomarker. Cancer Cell 39, 154–173 (2021).
Graf, R. P. et al. Comparative effectiveness of immune checkpoint inhibitors vs chemotherapy by tumor mutational burden in metastatic castration-resistant prostate cancer. JAMA Netw. Open 5, e225394 (2022).
Blades, R. A., Keating, P. J., McWilliam, L. J., George, N. J. & Stern, P. L. Loss of HLA class I expression in prostate cancer: implications for immunotherapy. Urology 46, 681–686 (1995).
Nesic, M. et al. The mutational profile of immune surveillance genes in diagnostic and refractory/relapsed DLBCLs. BMC Cancer 21, 829 (2021).
Luke, J. J., Bao, R., Sweis, R. F., Spranger, S. & Gajewski, T. F. WNT/beta-catenin pathway activation correlates with immune exclusion across human cancers. Clin. Cancer Res. 25, 3074–3083 (2019).
Datar, I. & Schalper, K. A. Epithelial-mesenchymal transition and immune evasion during lung cancer progression: the chicken or the egg? Clin. Cancer Res. 22, 3422–3424 (2016).
Ye, Z., Qian, Q., Jin, H. & Qian, Q. Cancer vaccine: learning lessons from immune checkpoint inhibitors. J. Cancer 9, 263–268 (2018).
Aragon-Ching, J. B., Williams, K. M. & Gulley, J. L. Impact of androgen-deprivation therapy on the immune system: implications for combination therapy of prostate cancer. Front. Biosci. 12, 4957–4971 (2007).
Bussard, K. M., Mutkus, L., Stumpf, K., Gomez-Manzano, C. & Marini, F. C. Tumor-associated stromal cells as key contributors to the tumor microenvironment. Breast Cancer Res. 18, 84 (2016).
Oh, S., Lee, J. H., Kwack, K. & Choi, S. W. Natural killer cell therapy: a new treatment paradigm for solid tumors. Cancers 11, 1534 (article number: 9) (2019).
Sebestyen, Z., Prinz, I., Dechanet-Merville, J., Silva-Santos, B. & Kuball, J. Translating gammadelta (gammadelta) T cells and their receptors into cancer cell therapies. Nat. Rev. Drug Discov. 19, 169–184 (2020).
Nicol, A. J. et al. Clinical evaluation of autologous gamma delta T cell-based immunotherapy for metastatic solid tumours. Br. J. Cancer 105, 778–786 (2011).
Zorko, N. A. & Ryan, C. J. Novel immune engagers and cellular therapies for metastatic castration-resistant prostate cancer: do we take a BiTe or ride BiKEs, TriKEs, and CARs? Prostate Cancer Prostatic Dis. 24, 986–996 (2021).
Kawazu, M. et al. HLA Class I analysis provides insight into the genetic and epigenetic background of immune evasion in colorectal cancer with high microsatellite instability. Gastroenterology 162, 799–812 (2022).
Jones, S. et al. Personalized genomic analyses for cancer mutation discovery and interpretation. Sci. Transl. Med. 7, 283ra253 (2015).
Li, H. & Durbin, R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 25, 1754–1760 (2009).
Van der Auwera, G. A. et al. From FastQ data to high confidence variant calls: the genome analysis Toolkit best practices pipeline. Curr. Protoc. Bioinforma. 43, 11 10 11–11 10 33 (2013).
Tate, J. G. et al. COSMIC: the catalogue of somatic mutations in cancer. Nucleic Acids Res. 47, D941–D947 (2019).
Antonarakis, E. S. et al. Clinical features and therapeutic outcomes in men with advanced prostate cancer and DNA mismatch repair gene mutations. Eur. Urol. 75, 378–382 (2019).
Sena, L. A. et al. Tumor frameshift mutation proportion predicts response to immunotherapy in mismatch repair-deficient prostate cancer. Oncologist 26, e270–e278 (2021).
Baller, J., Kono, T., Herman, A., & Zhang, Y. CHURP. PEARC19: Practice and Experience in Advanced Research Computing 2019: Rise of the Machines (Learning) 1–5 (Chicago, Illinois, 2019). https://doi.org/10.1145/3332186.3333156.
Wickham, H. ggplot2: Elegant Graphics for Data Analysis, (Springer, New York, NY, 2009).
Acknowledgements
J.R.L. is supported by NIH grant T32-GM008244. N.Z. is supported by a Prostate Cancer Foundation Young Investigator Award, The Randy Shaver Community Cancer Fund and a Department of Defense Early Investigator Research Award (W81XWH-22-1-0242). E.S.A. is partially funded by NIH grant P30-CA077598. The figures and workflows shown here were created with https://biorender.com/ and with ggplot256. The results shown here are in whole or part based upon data generated by the TCGA Research Network: https://www.cancer.gov/tcga. We also acknowledge the Broad Institute for use of its online databases, the DepMap portal, in the production of this work.
Author information
Authors and Affiliations
Contributions
H.E.B., L.A.S. and A.D. are co-first authors on this manuscript. H.E.B. contributed to the design of the project and the writing, edits, response to reviewers document, figures, and submission of the full manuscript. L.A.S. contributed to the conception of the project, the writing of the case report in the results section and abstract and provided insight on the clinical implications. A.D. contributed to the design of the project, design and implementation of all bioinformatics approaches, and wrote the corresponding data processing sections in the methods section. B.M. and C.D.M. contributed to the writing of the introduction. J.R.L. and N.Z. contributed to the conception of the revision documents. J.W. contributed to the insight regarding the bioinformatics analyses and pipelines. E.S. and C.G.D. participated in the design of the study. F.P.L. and L.M. contributed to the conception of the initial bioinformatics analyses presented in Figs. 1g and 2d. F.C.P. contributed to the quantification of CD8 and CD3 IHC staining in the manuscript and response to reviewers document as well as visualizations in Supplementary Fig. 1. T.L. contributed to the staining of B-cells, plasma cells, and tertiary lymphoid structures as described in the revision document. A.M.D.M. contributed to the staining and quantification of histology images in this case report including H&E, CD8, and CD3 in the manuscript and revision document. J.H. as the co-senior author on this manuscript contributed to the design of the project and the conceptualization of all analyses, and the writing, editing, and revisions of the final manuscript. E.S.A., as the corresponding author on this manuscript, contributed to the conception of the initial project, the writing, editing, and revisions and was the co-corresponding author on the initial clinical trial of which the patient was a part of. All authors contributed substantially to the design of the work, to the interpretation of the data, and to the editing and revising of the final manuscript document.
Corresponding author
Ethics declarations
Competing interests
L.A.S. has received research funding to her institution from Panbela Therapeutics. N.Z. is a consultant to Caris Life Sciences and Telix Pharmaceuticals. E.S.A. is a paid consultant to GT Biopharma, Guidepoint Global, FirstThought, GLG and CapitalOne Securities, and receives institutional research funding from MacroGenics, Inc. C.G.D. is a paid employee of Janssen Research and Development. J.H is a paid consultant/advisor to Caris Life Sciences and Astrin Biosciences. E.S.A. is a paid consultant/advisor to Janssen, Astellas, Sanofi, Dendreon, Pfizer, Amgen, Eli Lilly, Bayer, AstraZeneca, Bristol Myers Squibb, ESSA, Clovis, Merck, Curium, Blue Earth Diagnostics, Foundation Medicine, Exact Sciences and Invitae; has received research funding to his institution from Janssen, Johnson & Johnson, Sanofi, Dendreon, Genentech, Novartis, Tokai, Bristol Myers Squibb, Constellation, Bayer, AstraZeneca, Clovis and Merck; and is the coinventor of a patented AR-V7 biomarker technology that has been licensed to Qiagen. The other authors declare no competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Bergom, H.E., Sena, L.A., Day, A. et al. Divergent immune microenvironments in two tumor nodules from a patient with mismatch repair-deficient prostate cancer. npj Genom. Med. 9, 7 (2024). https://doi.org/10.1038/s41525-024-00392-1
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41525-024-00392-1
This article is cited by
-
Loss of heterozygosity impacts MHC expression on the immune microenvironment in CDK12-mutated prostate cancer
Molecular Cytogenetics (2024)