Metastatic site influences driver gene function in pancreatic cancer

Driver gene mutations can increase the metastatic potential of the primary tumor1-3, but their role in sustaining tumor growth at metastatic sites is poorly understood. A paradigm of such mutations is inactivation of SMAD4 - a transcriptional effector of TGFβ signaling - which is a hallmark of multiple gastrointestinal malignancies4,5. SMAD4 inactivation mediates TGFβ's remarkable anti- to pro-tumorigenic switch during cancer progression and can thus influence both tumor initiation and metastasis6-14. To determine whether metastatic tumors remain dependent on SMAD4 inactivation, we developed a mouse model of pancreatic ductal adenocarcinoma (PDAC) that enables Smad4 depletion in the pre-malignant pancreas and subsequent Smad4 reactivation in established metastases. As expected, Smad4 inactivation facilitated the formation of primary tumors that eventually colonized the liver and lungs. By contrast, Smad4 reactivation in metastatic disease had strikingly opposite effects depending on the tumor's organ of residence: suppression of liver metastases and promotion of lung metastases. Integrative multiomic analysis revealed organ-specific differences in the tumor cells' epigenomic state, whereby the liver and lungs harbored chromatin programs respectively dominated by the KLF and RUNX developmental transcription factors, with Klf4 depletion being sufficient to reverse Smad4's tumor-suppressive activity in liver metastases. Our results show how epigenetic states favored by the organ of residence can influence the function of driver genes in metastatic tumors. This organ-specific gene-chromatin interplay invites consideration of anatomical site in the interpretation of tumor genetics, with implications for the therapeutic targeting of metastatic disease.

Convergent evolution of BRCA2 reversion mutations under therapeutic pressure by PARP inhibition and platinum chemotherapy

Reversion mutations that restore wild-type function of the BRCA gene have been described as a key mechanism of resistance to Poly(ADP-ribose) polymerase (PARP) inhibitor therapy in BRCA-associated cancers. Here, we report a case of a patient with metastatic castration-resistant prostate cancer (mCRPC) with a germline BRCA2 mutation who developed acquired resistance to PARP inhibition. Extensive genomic interrogation of cell-free DNA (cfDNA) and tissue at baseline, post-progression, and postmortem revealed ten unique BRCA2 reversion mutations across ten sites. While several of the reversion mutations were private to a specific site, nine out of ten tumors contained at least one mutation, suggesting a powerful clonal selection for reversion mutations in the presence of therapeutic pressure by PARP inhibition. Variable cfDNA shed was seen across tumor sites, emphasizing a potential shortcoming of cfDNA monitoring for PARPi resistance. This report provides a genomic portrait of the temporal and spatial heterogeneity of prostate cancer under the selective pressure of a PARP inhibition and exposes limitations in the current strategies for detection of reversion mutations.

The nuclear factor ID3 endows macrophages with a potent anti-tumour activity

Macrophage activation is controlled by a balance between activating and inhibitory receptors1-7, which protect normal tissues from excessive damage during infection8,9 but promote tumour growth and metastasis in cancer7,10. Here we report that the Kupffer cell lineage-determining factor ID3 controls this balance and selectively endows Kupffer cells with the ability to phagocytose live tumour cells and orchestrate the recruitment, proliferation and activation of natural killer and CD8 T lymphoid effector cells in the liver to restrict the growth of a variety of tumours. ID3 shifts the macrophage inhibitory/activating receptor balance to promote the phagocytic and lymphoid response, at least in part by buffering the binding of the transcription factors ELK1 and E2A at the SIRPA locus. Furthermore, loss- and gain-of-function experiments demonstrate that ID3 is sufficient to confer this potent anti-tumour activity to mouse bone-marrow-derived macrophages and human induced pluripotent stem-cell-derived macrophages. Expression of ID3 is therefore necessary and sufficient to endow macrophages with the ability to form an efficient anti-tumour niche, which could be harnessed for cell therapy in cancer.

Islands of genomic stability in the face of genetically unstable metastatic cancer

Introduction

Metastatic cancer affects millions of people worldwide annually and is the leading cause of cancer-related deaths. Most patients with metastatic disease are not eligible for surgical resection, and current therapeutic regimens have varying success rates, some with 5-year survival rates below 5%. Here we test the hypothesis that metastatic cancer can be genetically targeted by exploiting single base substitution mutations unique to individual cells that occur as part of normal aging prior to transformation. These mutations are targetable because ~10% of them form novel tumor-specific "NGG" protospacer adjacent motif (PAM) sites targetable by CRISPR-Cas9.

Methods

Whole genome sequencing was performed on five rapid autopsy cases of patient-matched primary tumor, normal and metastatic tissue from pancreatic ductal adenocarcinoma decedents. CRISPR-Cas9 PAM targets were determined by bioinformatic tumor-normal subtraction for each patient and verified in metastatic samples by high-depth capture-based sequencing.

Results

We found that 90% of PAM targets were maintained between primary carcinomas and metastases overall. We identified rules that predict PAM loss or retention, where PAMs located in heterozygous regions in the primary tumor can be lost in metastases (private LOH), but PAMs occurring in regions of loss of heterozygosity (LOH) in the primary tumor were universally conserved in metastases.

Conclusions

Regions of truncal LOH are strongly retained in the presence of genetic instability, and therefore represent genetic vulnerabilities in pancreatic adenocarcinomas. A CRISPR-based gene therapy approach targeting these regions may be a novel way to genetically target metastatic cancer.

A microglia clonal inflammatory disorder in Alzheimer's Disease

Somatic genetic heterogeneity resulting from post-zygotic DNA mutations is widespread in human tissues and can cause diseases, however few studies have investigated its role in neurodegenerative processes such as Alzheimer's Disease (AD). Here we report the selective enrichment of microglia clones carrying pathogenic variants, that are not present in neuronal, glia/stromal cells, or blood, from patients with AD in comparison to age-matched controls. Notably, microglia-specific AD-associated variants preferentially target the MAPK pathway, including recurrent CBL ring-domain mutations. These variants activate ERK and drive a microglia transcriptional program characterized by a strong neuro-inflammatory response, both in vitro and in patients. Although the natural history of AD-associated microglial clones is difficult to establish in human, microglial expression of a MAPK pathway activating variant was previously shown to cause neurodegeneration in mice, suggesting that AD-associated neuroinflammatory microglial clones may contribute to the neurodegenerative process in patients.

ConDoR: tumor phylogeny inference with a copy-number constrained mutation loss model

A tumor contains a diverse collection of somatic mutations that reflect its past evolutionary history and that range in scale from single nucleotide variants (SNVs) to large-scale copy-number aberrations (CNAs). However, no current single-cell DNA sequencing (scDNA-seq) technology produces accurate measurements of both SNVs and CNAs, complicating the inference of tumor phylogenies. We introduce a new evolutionary model, the constrained k-Dollo model, that uses SNVs as phylogenetic markers but constrains losses of SNVs according to clusters of cells. We derive an algorithm, ConDoR, that infers phylogenies from targeted scDNA-seq data using this model. We demonstrate the advantages of ConDoR on simulated and real scDNA-seq data.

A targetable pathway to eliminate TRA-1-60+/TRA-1-81+ chemoresistant cancer cells

Chemoresistance is a primary cause of treatment failure in pancreatic cancer. Identifying cell surface markers specifically expressed in chemoresistant cancer cells (CCCs) could facilitate targeted therapies to overcome chemoresistance. We performed an antibody-based screen and found that TRA-1-60 and TRA-1-81, two 'stemness' cell surface markers, are highly enriched in CCCs. Furthermore, TRA-1-60+/TRA-1-81+ cells are chemoresistant compared to TRA-1-60-/TRA-1-81- cells. Transcriptome profiling identified UGT1A10, shown to be both necessary and sufficient to maintain TRA-1-60/TRA-1-81 expression and chemoresistance. From a high-content chemical screen, we identified Cymarin, which downregulates UGT1A10, eliminates TRA-1-60/TRA-1-81 expression, and increases chemosensitivity both in vitro and in vivo. Finally, TRA-1-60/TRA-1-81 expression is highly specific in primary cancer tissue and positively correlated with chemoresistance and short survival, which highlights their potentiality for targeted therapy. Therefore, we discovered a novel CCC surface marker regulated by a pathway that promotes chemoresistance, as well as a leading drug candidate to target this pathway.

Subclonal Somatic Copy-Number Alterations Emerge and Dominate in Recurrent Osteosarcoma

Multiple large-scale genomic profiling efforts have been undertaken in osteosarcoma to define the genomic drivers of tumorigenesis, therapeutic response, and disease recurrence. The spatial and temporal intratumor heterogeneity could also play a role in promoting tumor growth and treatment resistance. We conducted longitudinal whole-genome sequencing of 37 tumor samples from 8 patients with relapsed or refractory osteosarcoma. Each patient had at least one sample from a primary site and a metastatic or relapse site. Subclonal copy-number alterations were identified in all patients except one. In 5 patients, subclones from the primary tumor emerged and dominated at subsequent relapses. MYC gain/amplification was enriched in the treatment-resistant clones in 6 of 7 patients with multiple clones. Amplifications in other potential driver genes, such as CCNE1, RAD21, VEGFA, and IGF1R, were also observed in the resistant copy-number clones. A chromosomal duplication timing analysis revealed that complex genomic rearrangements typically occurred prior to diagnosis, supporting a macroevolutionary model of evolution, where a large number of genomic aberrations are acquired over a short period of time followed by clonal selection, as opposed to ongoing evolution. A mutational signature analysis of recurrent tumors revealed that homologous repair deficiency (HRD)-related SBS3 increases at each time point in patients with recurrent disease, suggesting that HRD continues to be an active mutagenic process after diagnosis. Overall, by examining the clonal relationships between temporally and spatially separated samples from patients with relapsed/refractory osteosarcoma, this study sheds light on the intratumor heterogeneity and potential drivers of treatment resistance in this disease.

SIGNIFICANCE

The chemoresistant population in recurrent osteosarcoma is subclonal at diagnosis, emerges at the time of primary resection due to selective pressure from neoadjuvant chemotherapy, and is characterized by unique oncogenic amplifications.

Tissue-specific features of the T cell repertoire after allogeneic hematopoietic cell transplantation in human and mouse

T cells are the central drivers of many inflammatory diseases, but the repertoire of tissue-resident T cells at sites of pathology in human organs remains poorly understood. We examined the site-specificity of T cell receptor (TCR) repertoires across tissues (5 to 18 tissues per patient) in prospectively collected autopsies of patients with and without graft-versus-host disease (GVHD), a potentially lethal tissue-targeting complication of allogeneic hematopoietic cell transplantation, and in mouse models of GVHD. Anatomic similarity between tissues was a key determinant of TCR repertoire composition within patients, independent of disease or transplant status. The T cells recovered from peripheral blood and spleens in patients and mice captured a limited portion of the TCR repertoire detected in tissues. Whereas few T cell clones were shared across patients, motif-based clustering revealed shared repertoire signatures across patients in a tissue-specific fashion. T cells at disease sites had a tissue-resident phenotype and were of donor origin based on single-cell chimerism analysis. These data demonstrate the complex composition of T cell populations that persist in human tissues at the end stage of an inflammatory disorder after lymphocyte-directed therapy. These findings also underscore the importance of studying T cell in tissues rather than blood for tissue-based pathologies and suggest the tissue-specific nature of both the endogenous and posttransplant T cell landscape.

Cancer cells co-evolve with retrotransposons to mitigate viral mimicry

Overexpression of repetitive elements is an emerging hallmark of human cancers ¹ . Diverse repeats can mimic viruses by replicating within the cancer genome through retrotransposition, or presenting pathogen-associated molecular patterns (PAMPs) to the pattern recognition receptors (PRRs) of the innate immune system ^2-5 . Yet, how specific repeats affect tumor evolution and shape the tumor immune microenvironment (TME) in a pro- or anti-tumorigenic manner remains poorly defined. Here, we integrate whole genome and total transcriptome data from a unique autopsy cohort of multiregional samples collected in pancreatic ductal adenocarcinoma (PDAC) patients, into a comprehensive evolutionary analysis. We find that more recently evolved S hort I nterspersed N uclear E lements (SINE), a family of retrotransposable repeats, are more likely to form immunostimulatory double-strand RNAs (dsRNAs). Consequently, younger SINEs are strongly co-regulated with RIG-I like receptor associated type-I interferon genes but anti-correlated with pro-tumorigenic macrophage infiltration. We discover that immunostimulatory SINE expression in tumors is regulated by either L ong I nterspersed N uclear E lements 1 (LINE1/L1) mobility or ADAR1 activity in a TP53 mutation dependent manner. Moreover, L1 retrotransposition activity tracks with tumor evolution and is associated with TP53 mutation status. Altogether, our results suggest pancreatic tumors actively evolve to modulate immunogenic SINE stress and induce pro-tumorigenic inflammation. Our integrative, evolutionary analysis therefore illustrates, for the first time, how dark matter genomic repeats enable tumors to co-evolve with the TME by actively regulating viral mimicry to their selective advantage.

Clonal evolution during metastatic spread in high-risk neuroblastoma

Patients with high-risk neuroblastoma generally present with widely metastatic disease and often relapse despite intensive therapy. As most studies to date focused on diagnosis-relapse pairs, our understanding of the genetic and clonal dynamics of metastatic spread and disease progression remain limited. Here, using genomic profiling of 470 sequential and spatially separated samples from 283 patients, we characterize subtype-specific genetic evolutionary trajectories from diagnosis through progression and end-stage metastatic disease. Clonal tracing timed disease initiation to embryogenesis. Continuous acquisition of structural variants at disease-defining loci (MYCN, TERT, MDM2-CDK4) followed by convergent evolution of mutations targeting shared pathways emerged as the predominant feature of progression. At diagnosis metastatic clones were already established at distant sites where they could stay dormant, only to cause relapses years later and spread via metastasis-to-metastasis and polyclonal seeding after therapy.

Abstract 1169: Genomic evolution of pancreatic cancer at single-cell resolution

As the breadth of genomic sequencing datasets increases, we can engage more directly with the evolutionary principles governing cancer progression. But as clonal evolution happens at the single-cell level, one must rely on strong assumptions when making conclusions with bulk-sequencing results, which fails in capturing clonal heterogeneity to its fullest.

Herein we have developed and optimized a high-throughput, high-depth targeted single-nucleus DNA-seq (snDNA-seq) technique (doi.org/10.1101/2022.03.06.483206) for archival primary solid tumor samples. We focused on pancreatic ductal adenocarcinoma (PDAC), well known as one of the most lethal cancers, and sequenced over 200,000 single nuclei from 80 archival primary samples of 25 PDAC patients. The samples included both early- and late-stage diagnoses and multiregional sampling from primary tumors and metastasis to capture clonal heterogeneity on both temporal and spatial scales.

With significant increase in sensitivity than bulk (down to mutations in 0.1% cells), we discovered thousands of novel mutations per sample on our 120,000 base-pair-long panel regions, suggesting a mutation rate higher than previously estimated. A small fraction of these mutations is in 1-10% single cells and are enriched in early-stage samples. They form mutually exclusive clones which functionally target key pathways including TGF-β, homologous recombination, suggesting subclonal evolution under positive selection at the early stage of cancer. Most novel mutations are in <1% single cells sampled and enabled us to measure convergent evolution and positive/negative selection within each tumor.

We next revisited PDAC’s genomic evolution model established by bulk studies. It posits that PDAC often arises when KRAS hotspot mutation-bearing precursors acquire TP53 and/or CDKN2A inactivation through stepwise and punctuated evolution. While it assumes that TGF-β inactivating mutations are present in all cancer cells (clonal), we found that they are targeted in a highly subclonal manner; moreover, short mutations and focal copy number variations occur in a stepwise manner over time leading to the most “fit” genotype. In many PDACs whose bulk results show no alteration to the TGF-β pathway, snDNA-seq shows focal deletions that are likely below bulk’s sensitivity.

Ongoing studies have begun to extend these analyses to longitudinal samples to study treatment response; normal pancreas tissues to study pancreas cells’ clonal evolution in aging and chronic disease conditions; blood samples to investigate circulating tumor cells in PDAC patients. Computational pipelines and analysis tools are being built as platform for more in-depth analyses. Overall, the high-throughput snDNA-seq technique brings genomic study of PDAC to a much higher resolution and holds the promise to not only inform precision medicine but also shed light on many fundamental questions on cancer evolution.

Citation Format: Haochen Zhang, Palash Sashittal, Elias-Ramzey Karnoub, Benjamin J. Raphael, Christine A. Iacobuzio-Donahue. Genomic evolution of pancreatic cancer at single-cell resolution [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 1169.

Abstract 2148: Robust detection of somatic genetic alterations in pancreatic cancer ascites

Introduction: Regardless of the stage at diagnosis most patients with pancreatic ductal adenocarcinoma develop peritoneal disease and some malignant ascites (MA) as well. Prior studies have shown that MA negatively affects overall treatment efficacy and survival. Despite the clinical significance of MA it has not been studied to any great extent.

Methods: We collected MA and matched normal tissue samples at autopsy from 20 PDAC patients who were initially diagnosed at stages IIB to IV. Whole exome or targeted sequencing was previously performed on each PDAC. Each MA sample was centrifuged twice at 4000 RPM first and then at 15000 RPM to separate the cell pellet (CP) from the cell-free ascites fluid. We next extracted DNA from the CP, matched normal tissue, and the cell-free DNA (cfDNA) from the ascites fluid, and all were submitted to the Genomics Core for MSK-IMPACT, a targeted cancer gene panel representing 505 genes.

Results: Results of the first five patients are complete and the remaining are in process. Comparison of the CPs and/or cfDNA to the matched tumor samples indicated 100% concordance for detected variants. However, the somatic alterations of the CP specifically versus the matched cfDNA were divergent in all patients analyzed thus far. Virtually all copy number alterations in all patients were deep deletions (range 66 to 187 cancer genes deleted) affecting multiple DNA repair pathways including homologous recombination deficiency and microsatellite repair.

Conclusions: Samples of MA, when both the cell pellet and cfDNA are sequenced, accurately represent the genetic features of the matched PDAC tissue and may serve as an alternative mode of sampling for precision medicine. Differences in the genetics of the CP versus the cfDNA suggest polyclonality in the peritoneal space. Moreover, the finding of deep deletions in targetable DNA repair pathways suggest a therapeutic vulnerability for exploration. Given that paracentesis is often performed in the palliative setting and may be performed multiple times over the course of a patients’ management, it also offers an opportunity to determine how clonal dynamics in the peritoneal space change over time. Patients with MA have poor overall survival compared to patients without MA so these patients may benefit from this type of tracking which could potentially help with their treatment.

Citation Format: Rajya L. Kappagantula, Alvin P. Makohon-Moore, Shigeaki Umeda, Elias-Ramzey R. Karnoub, Jerry P. Melchor, Laura D. Wood, Christine A. Iacobuzio-Donahue. Robust detection of somatic genetic alterations in pancreatic cancer ascites [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 2148.

Genomic Biomarkers Associated with Response to Induction Chemotherapy in Patients with Localized Pancreatic Ductal Adenocarcinoma

PURPOSE

There is increasing use of neoadjuvant chemotherapy in the management of localized pancreatic ductal adenocarcinoma (PDAC), yet there are few validated biomarkers to guide therapy selection. We aimed to determine whether somatic genomic biomarkers predict response to induction FOLFIRINOX or gemcitabine/nab-paclitaxel.

EXPERIMENTAL DESIGN

This single-institution cohort study included consecutive patients (N = 322) with localized PDAC (2011-2020) who received at least one cycle of FOLFIRINOX (N = 271) or gemcitabine/nab-paclitaxel (N = 51) as initial treatment. We assessed somatic alterations in four driver genes (KRAS, TP53, CDKN2A, and SMAD4) by targeted next-generation sequencing, and determined associations between these alterations and (1) rate of metastatic progression during induction chemotherapy, (2) surgical resection, and (3) complete/major pathologic response.

RESULTS

The alteration rates in driver genes KRAS, TP53, CDKN2A, and SMAD4 were 87.0%, 65.5%, 26.7%, and 19.9%, respectively. For patients receiving first-line FOLFIRINOX, SMAD4 alterations were uniquely associated with metastatic progression (30.0% vs. 14.5%; P = 0.009) and decreased rate of surgical resection (37.1% vs. 66.7%; P < 0.001). For patients receiving induction gemcitabine/nab-paclitaxel, alterations in SMAD4 were not associated with metastatic progression (14.3% vs. 16.2%; P = 0.866) nor decreased rate of surgical resection (33.3% vs. 41.9%; P = 0.605). Major pathologic response was rare (6.3%) and not associated with type of chemotherapy regimen.

CONCLUSIONS

SMAD4 alterations were associated with more frequent development of metastasis and lower probability of reaching surgical resection during neoadjuvant FOLFIRINOX but not gemcitabine/nab-paclitaxel. Confirmation in a larger, diverse patient cohort will be important before prospective evaluation of SMAD4 as a genomic biomarker to guide treatment selection.

STING inhibits the reactivation of dormant metastasis in lung adenocarcinoma

Metastasis frequently develops from disseminated cancer cells that remain dormant after the apparently successful treatment of a primary tumour. These cells fluctuate between an immune-evasive quiescent state and a proliferative state liable to immune-mediated elimination1-6. Little is known about the clearing of reawakened metastatic cells and how this process could be therapeutically activated to eliminate residual disease in patients. Here we use models of indolent lung adenocarcinoma metastasis to identify cancer cell-intrinsic determinants of immune reactivity during exit from dormancy. Genetic screens of tumour-intrinsic immune regulators identified the stimulator of interferon genes (STING) pathway as a suppressor of metastatic outbreak. STING activity increases in metastatic progenitors that re-enter the cell cycle and is dampened by hypermethylation of the STING promoter and enhancer in breakthrough metastases or by chromatin repression in cells re-entering dormancy in response to TGFβ. STING expression in cancer cells derived from spontaneous metastases suppresses their outgrowth. Systemic treatment of mice with STING agonists eliminates dormant metastasis and prevents spontaneous outbreaks in a T cell- and natural killer cell-dependent manner-these effects require cancer cell STING function. Thus, STING provides a checkpoint against the progression of dormant metastasis and a therapeutically actionable strategy for the prevention of disease relapse.

Application of high-throughput single-nucleus DNA sequencing in pancreatic cancer

Despite insights gained by bulk DNA sequencing of cancer it remains challenging to resolve the admixture of normal and tumor cells, and/or of distinct tumor subclones; high-throughput single-cell DNA sequencing circumvents these and brings cancer genomic studies to higher resolution. However, its application has been limited to liquid tumors or a small batch of solid tumors, mainly because of the lack of a scalable workflow to process solid tumor samples. Here we optimize a highly automated nuclei extraction workflow that achieves fast and reliable targeted single-nucleus DNA library preparation of 38 samples from 16 pancreatic ductal adenocarcinoma patients, with an average library yield per sample of 2867 single nuclei. We demonstrate that this workflow not only performs well using low cellularity or low tumor purity samples but reveals genomic evolution patterns of pancreatic ductal adenocarcinoma as well.

Genetic heterogeneity of intrahepatic cholangiocarcinoma: Implications for outcome

595

Background: Intrahepatic cholangiocarcinoma (IHC) is characterized by marked clinical heterogeneity, likely the result of multiple cells of origin and variable driver gene alterations. The hidden-genome classifier is a statistical algorithm that classifies tumors by integrating multi-level genomic features. In this study, we trained the hidden-genome classifier with extrahepatic cholangiocarcinoma (EHC), gallbladder cancer (GBC) and hepatocellular carcinoma (HCC) as extremes of a spectrum to quantify the genetic heterogeneity of IHC with a view toward improved tumor classification. Methods: An IRB approved retrospective review of patients with biopsy confirmed IHC, EHC, GBC and HCC was conducted. All tumors were subjected to MSK-IMPACT to determine the mutational profile. A two-class model was built and internally validated with the genomic data of EHC/GBC as one class and HCC as the other class. IHC tumors were analyzed in the model and classified into three groups based on their proportional genetic resemblance to EHC/GBC (Biliary Class) or HCC (HCC Class), with the remainder as Intermediate Class. The classification thresholds were 90% resemblance to EHC/GBC or HCC and were determined by the inflection point of predicted survival. The survivals of the three groups were analyzed and compared. Results: A total of 1497 patients were included: IHC (733), EHC (208), GBC (258) and HCC (298). 527 IHC tumors with complete metagenetic information were analyzed in the model, showing a continuous spectrum of alterations, ranging from Biliary Class (122 tumors), Intermediate Class (375 tumors) to HCC Class (30 tumor). The biliary-class IHC was characterized by frequent alterations of IDH1 R132C, KRAS, SMAD4, ERBB2 gain, MDM2 gain, and CKDN2A loss, while the HCC-class IHC was primarily characterized by TERT alterations. In patients with unresected IHCs, the median survival ranged from 1 year (CI 0.77, 1.5) in Biliary Class, 1.8 years (CI 1.5, 2.0) in Intermediate Class, to 2 years (CI 0.93, NR) in HCC Class. In patients subjected to resection, the median survival of Biliary Class (2.4 years, CI 2.1, NR) was lower than both the Intermediate Class (5.1 years, CI 4.8, 6.9) and the HCC Class (3.4 years, CI 2.7, NR). Conclusions: By integrating multi-level genomic features, we leveraged the mutational heterogeneity to classify IHC based on its resemblance to EHC/GBC or HCC tumors. We found that the survival in IHC patients appeared to decline with increasing genomic similarity to Biliary Class. The results support a genomic basis for IHC’s variable clinical behavior and point to a role of mutational testing to guide clinical intervention. [Table: see text]

Concordance in oncogenic alterations between primary and recurrent/metastatic cholangiocarcinoma pairs using targeted next-generation sequencing

604

Background: The genetic background of cholangiocarcinoma (CCA) commonly involves alterations in kinase signaling, tumor suppression, oxidative stress modulation, and proto-oncogenic coupling pathways. Novel agents targeting such pathways have shown promise in systemic treatment; however, studies examining differences in the mutational landscapes between primary and recurrent, metastatic, or progressive disease after systemic therapy are lacking. The present study aimed to determine if recurrent, metastatic, or progressive disease genetically parallels the primary or not. Methods: Patients with biopsy proven CCA (primary tumor and paired recurrent/metastatic or progressive disease) from two institutions (MSKCC and Duke) were identified. Targeted next-generation sequencing (Integrated Mutation Profiling of Actionable Cancer Targets (IMPACT)) capturing single nucleotide variants, copy number alterations, and structural variants was used to compare driver alteration concordance across the paired samples. Subgroup analyses were performed based on exposure to systemic therapy in patients with disease progression and tumor type (intrahepatic versus extrahepatic). Results: Sample pairs from 65 patients with intrahepatic (ICCA, n=54) and extrahepatic CCA (ECCA, n=11) were analyzed. Median time between samples was 19.6 months (range 2.7 - 122.9). Some de novo alterations were identified in recurrent/metastatic samples, but overall concordance (70%) was demonstrated between patient pairs for common oncogenic driver genes (Table). Subgroup analyses of summative ICCA and ECCA mutations revealed concordance of 65% and 88%, respectively. Concordance was also demonstrated between pairs exposed to systemic therapy between sample collections (n=50, 71%). Conclusions: In this dataset of CCA patients, a concordance rate of 70% was identified in the genomic alterations between primary and recurrent/metastatic pairs, and this did not appear to be altered by prior treatment with systemic chemotherapy. While limited by sample size, concordance in ICCA pairs was lower than that seen in ECCA. [Table: see text]

Subclonal somatic copy number alterations emerge and dominate in recurrent osteosarcoma

Multiple large-scale tumor genomic profiling efforts have been undertaken in osteosarcoma, however, little is known about the spatial and temporal intratumor heterogeneity and how it may drive treatment resistance. We performed whole-genome sequencing of 37 tumor samples from eight patients with relapsed or refractory osteosarcoma. Each patient had at least one sample from a primary site and a metastatic or relapse site. We identified subclonal copy number alterations in all but one patient. We observed that in five patients, a subclonal copy number clone from the primary tumor emerged and dominated at subsequent relapses. MYC gain/amplification was enriched in the treatment-resistant clone in 6 out of 7 patients with more than one clone. Amplifications in other potential driver genes, such as CCNE1, RAD21, VEGFA, and IGF1R, were also observed in the resistant copy number clones. Our study sheds light on intratumor heterogeneity and the potential drivers of treatment resistance in osteosarcoma.

ConDoR: Tumor phylogeny inference with a copy-number constrained mutation loss model

Tumors consist of subpopulations of cells that harbor distinct collections of somatic mutations. These mutations range in scale from single nucleotide variants (SNVs) to large-scale copy-number aberrations (CNAs). While many approaches infer tumor phylogenies using SNVs as phylogenetic markers, CNAs that overlap SNVs may lead to erroneous phylogenetic inference. Specifically, an SNV may be lost in a cell due to a deletion of the genomic segment containing the SNV. Unfortunately, no current single-cell DNA sequencing (scDNA-seq) technology produces accurate measurements of both SNVs and CNAs. For instance, recent targeted scDNA-seq technologies, such as Mission Bio Tapestri, measure SNVs with high fidelity in individual cells, but yield much less reliable measurements of CNAs. We introduce a new evolutionary model, the constrained k-Dollo model, that uses SNVs as phylogenetic markers and partial information about CNAs in the form of clustering of cells with similar copy-number profiles. This copy-number clustering constrains where loss of SNVs can occur in the phylogeny. We develop ConDoR (Constrained Dollo Reconstruction), an algorithm to infer tumor phylogenies from targeted scDNA-seq data using the constrained k-Dollo model. We show that ConDoR outperforms existing methods on simulated data. We use ConDoR to analyze a new multi-region targeted scDNA-seq dataset of 2153 cells from a pancreatic ductal adenocarcinoma (PDAC) tumor and produce a more plausible phylogeny compared to existing methods that conforms to histological results for the tumor from a previous study. We also analyze a metastatic colorectal cancer dataset, deriving a more parsimonious phylogeny than previously published analyses and with a simpler monoclonal origin of metastasis compared to the original study.

Code availability

Software is available at https://github.com/raphael-group/constrained-Dollo.

Abstract A055: Altered RNA splicing causes pancreatic cancer and exposes a therapeutic vulnerability

With the aim to identify new second-hit mutations that cause pancreatic ductal adenocarcinoma (PDAC), other than mutations in p53, we performed a genetic interaction analysis based on the concept of mutually exclusive mutations, which suggests that mutations that are exclusive to one another function within the same pathway. Using our unbiased quantitative methods to evaluate the effects of somatic mutations on cancer we analyzed &gt;3000 PDAC cases to prioritize mutations among hundreds that appear at low frequencies (5-10%) and that are mutually exclusive with mutations in p53. We found that mutations in RNA splicing factors SF3B1 and RBM10 (~15% of cases), were among the most significant, and mutually exclusive, to mutant p53. Thus, we hypothesized that aberrant RNA splicing promoted by SF3B1K700E and loss of RBM10 cause an oncogenic program that leads to tumorigenesis and therapy resistance. To test this hypothesis, first, we engineered mouse models to co-express KrasG12D with either Sf3b1K700E or loss of Rbm10 in Mist1 and Pdx1 driven expressing pancreatic cells. Interestingly, mice harboring KrasG12D and Sf3b1 K700E or loss of RBM10—like mutant p53—caused PDACs in mice. To identify the RNA splicing mechanism that leads to PDAC formation, we performed deep RNA sequencing and unbiased transcriptome-wide splicing analyses in isolated PDAC cells bearing wild-type, SF3B1K700E and loss of RBM10, derived from murine and patient PDACs. We found that like the splicing defects induced by p53mut—SF3B1K700E or loss of RBM10—induce splicing defects in sequence-specific coding exons in mRNAs derived from chromosomes 19, 7, 1 and 2 and impacting mainly GTPase activation and lipid metabolism pathways across species models. Consistent with this common mechanism, we found that PDAC cells do not tolerate the forced expression of mutant p53 with either SF3B1K700E or loss of Rbm10, supporting the mutual exclusivity findings from patient samples. To determine a personalized therapy pathway for patients bearing mutant splicing-factor tumors, we found that PDAC cells with SF3B1K700E or RBM10 loss cells are more sensitive to Gemcitabine rather than 5FU, which are the two main chemotherapeutic components of the two standard-of-care therapies. In addition, these mutations sensitize cells to small-molecule splicing inhibitor H3B-8800, which has advanced to phase 2 trial for hematologic malignancies. Lastly, we found that combination of Gemcitabine and H3B-8800 synergize to selective kill mutant splicing cells. We are currently testing these combination therapies in our genetically engineered mouse models and patient derived organoids. We are starting a phase 2 trial for Gemcitabine/Abraxane with escalating doses of H3B-8800 for tumors that have either SF3B1K700E, RBM10 loss or neomorphic p53, all which have aberrant RNA splicing. This work has determined novel drivers and mechanisms of PDAC development and a precision therapeutic strategy for the treatment of 45% pancreatic cancer patients (SF3B1K700E [5%], RBM10 loss [10%] and neomorphic p53 [30%]).

Citation Format: Natasha Pinto Medici, Diana Martinez Saucedo, Danny Lee, Li-Ting Ku, Robert Tseng, Vincent Cannataro, Deanne Yugawa, Sumedha Chowdhury, Jeffrey Townsend, Christine A. Iacobuzio-Donahue, Omar Abdel-Wahab, Steven D. Leach, Luisa Escobar-Hoyos. Altered RNA splicing causes pancreatic cancer and exposes a therapeutic vulnerability [abstract]. In: Proceedings of the AACR Special Conference on Pancreatic Cancer; 2022 Sep 13-16; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2022;82(22 Suppl):Abstract nr A055.

MACHETE identifies interferon-encompassing chromosome 9p21.3 deletions as mediators of immune evasion and metastasis

The most prominent homozygous deletions in cancer affect chromosome 9p21.3 and eliminate CDKN2A/B tumor suppressors, disabling a cell-intrinsic barrier to tumorigenesis. Half of 9p21.3 deletions, however, also encompass a type I interferon (IFN) gene cluster; the consequences of this co-deletion remain unexplored. To functionally dissect 9p21.3 and other large genomic deletions, we developed a flexible deletion engineering strategy, MACHETE (molecular alteration of chromosomes with engineered tandem elements). Applying MACHETE to a syngeneic mouse model of pancreatic cancer, we found that co-deletion of the IFN cluster promoted immune evasion, metastasis and immunotherapy resistance. Mechanistically, IFN co-deletion disrupted type I IFN signaling in the tumor microenvironment, leading to marked changes in infiltrating immune cells and escape from CD8+ T-cell surveillance, effects largely driven by the poorly understood interferon epsilon. These results reveal a chromosomal deletion that disables both cell-intrinsic and cell-extrinsic tumor suppression and provide a framework for interrogating large deletions in cancer and beyond.

BRCA mutational status shapes the stromal microenvironment of pancreatic cancer linking clusterin expression in cancer associated fibroblasts with HSF1 signaling

Tumors initiate by mutations in cancer cells, and progress through interactions of the cancer cells with non-malignant cells of the tumor microenvironment. Major players in the tumor microenvironment are cancer-associated fibroblasts (CAFs), which support tumor malignancy, and comprise up to 90% of the tumor mass in pancreatic cancer. CAFs are transcriptionally rewired by cancer cells. Whether this rewiring is differentially affected by different mutations in cancer cells is largely unknown. Here we address this question by dissecting the stromal landscape of BRCA-mutated and BRCA Wild-type pancreatic ductal adenocarcinoma. We comprehensively analyze pancreatic cancer samples from 42 patients, revealing different CAF subtype compositions in germline BRCA-mutated vs. BRCA Wild-type tumors. In particular, we detect an increase in a subset of immune-regulatory clusterin-positive CAFs in BRCA-mutated tumors. Using cancer organoids and mouse models we show that this process is mediated through activation of heat-shock factor 1, the transcriptional regulator of clusterin. Our findings unravel a dimension of stromal heterogeneity influenced by germline mutations in cancer cells, with direct implications for clinical research.

Systematic Comparison of Pancreatic Ductal Adenocarcinoma Models Identifies a Conserved Highly Plastic Basal Cell State

Intratumoral heterogeneity and cellular plasticity have emerged as hallmarks of cancer, including pancreatic ductal adenocarcinoma (PDAC). As PDAC portends a dire prognosis, a better understanding of the mechanisms underpinning cellular diversity in PDAC is crucial. Here, we investigated the cellular heterogeneity of PDAC cancer cells across a range of in vitro and in vivo growth conditions using single-cell genomics. Heterogeneity contracted significantly in two-dimensional and three-dimensional cell culture models but was restored upon orthotopic transplantation. Orthotopic transplants reproducibly acquired cell states identified in autochthonous PDAC tumors, including a basal state exhibiting coexpression and coaccessibility of epithelial and mesenchymal genes. Lineage tracing combined with single-cell transcriptomics revealed that basal cells display high plasticity in situ. This work defines the impact of cellular growth conditions on phenotypic diversity and uncovers a highly plastic cell state with the capacity to facilitate state transitions and promote intratumoral heterogeneity in PDAC.

SIGNIFICANCE

This work provides important insights into how different model systems of pancreatic ductal adenocarcinoma mold the phenotypic space of cancer cells, highlighting the power of in vivo models.

Abstract B022: Subclonal somatic copy number alterations emerge and dominate in relapsed/refractory osteosarcoma

Objective: Multiple large-scale tumor genomic profiling efforts have been undertaken in osteosarcoma, however little is known about the spatial and temporal intratumor heterogeneity and how it may drive treatment resistance. Methods: We performed 30-80x whole genome sequencing (WGS) of 37 tumor samples from 8 patients with relapsed or refractory osteosarcoma. Each patient had at least one sample from a primary site and one sample from a metastatic or relapse site. A set of high confidence single nucleotide variants (SNV), copy number alterations (CNA), structural variations (SV) were called for each sample using our pediatric expanded genomics pipeline and an evolutionary analysis was performed using a custom pipeline of computational tools. Results: Of the 8 patients in our cohort, 4 had localized disease at diagnosis (OSCE4, OSCE5, OSCE6, OSCE9) and 4 had metastatic disease at diagnosis (OSCE1, OSCE2, OSCE3, OSCE10). There were 17 samples from primary sites, 7 were pretreatment biopsies, 10 from on therapy primary resections. 20 samples came from metastatic sites, 15 of which were from lung metastases. Driver gene SNV’s were identified in 5 of 8 patients, including TP53 (OSCE1), ATRX (OSCE3, OSCE10), RB1 (OSCE4), and CDKN2A (OSCE9). There were no new driver SNV’s that emerged post-therapy in any patient. HATCHet, an algorithm that infers clone-specific copy number alterations, identified subclonal CNAs in all but one patient (OSCE2). In the 7 patients with subclonal CNAs, 6 had two copy number clones identified, and 1 patient (OSCE10) had three copy number clones identified. In 5 patients (OSCE1, OSCE4, OSCE5, OSCE6, OSCE10) there is a copy number clone that is subclonal in the primary tumor which emerges and dominates at subsequent relapses. The resistant clone in each of these cases had either MYC gain/amplification. Amplifications in CCNE1 (OSCE1), RAD21 (OSCE4, OSCE5, OSCE10), VEGFA (OSCE1, OSCE9), IGF1R (OSCE6) were also identified as potential drivers in the resistant copy number clones. In two of these patients (OSCE1, OSCE6), this treatment-resistant subclone becomes the dominant copy number clone by the time of primary resection. SNV based phylogenies revealed a heterogenous mix of monoclonal and polyclonal seeding of metastases and monophyletic and polyphyletic modes of dissemination. Over half the new mutations acquired in recurrent disease were attributed to HRD or cisplatin mutational signatures. TP53 structural variants were seen in 6/8 patients (OSCE2, OSCE3, OSCE4, OSCE6, OSCE9, OSCE10). New structural variants involving driver genes were only detected in one relapse sample from patient OSCE10 (DMD deletion). Conclusion: Subclonal copy number clones emerge and dominate in relapsed osteosarcoma, with MYC gain/amplification a defining characteristic in our cohort. Selective pressure from neoadjuvant chemotherapy reveals this clone at the time of primary resection, highlighting that genomic profiling at this time point may be more reflective of its metastatic potential.

Citation Format: Michael D. Kinnaman, Simone Zaccaria, Alvin Makohon-Moore, Gunes Gundem, Juan E. Arango Ossa, Nancy Bouvier, Filemon S. Dela Cruz, Meera Hameed, Julia Lynne Glade Bender, William D. Tap, Paul Meyers, Elli Papaemmanuil, Andrew Kung, Christine A. Iacobuzio-Donahue. Subclonal somatic copy number alterations emerge and dominate in relapsed/refractory osteosarcoma [abstract]. In: Proceedings of the AACR Special Conference: Sarcomas; 2022 May 9-12; Montreal, QC, Canada. Philadelphia (PA): AACR; Clin Cancer Res 2022;28(18_Suppl):Abstract nr B022.

Abstract 704: Development of a patient-derived xenograft (PDX) modeling program to enable pediatric precision medicine

Background: Recapitulation of the full spectrum of genomic changes driving patient tumors have resulted in increased use of patient-derived xenograft (PDX) models in studies of basic cancer biology and preclinical drug development. Given the translational potential of PDXs and limited availability of pediatric cancer models, we established a PDX program to expand the existing collection of pediatric PDXs in the community and enable pre- and post-clinical studies.

Methods: PDX generation requests were integrated into clinical workflows to maximize identification of eligible patients for informed consent and tissue collection at Memorial Sloan Kettering Cancer Center. Methodologies for tissue procurement and cryopreservation were optimized to facilitate implantation into host immunodeficient mice and enable multi-institutional tissue exchange for model building. A bioinformatics pipeline was established to allow molecular validation of engrafted PDXs using a next-generation targeted gene panel (MSK-IMPACT) evaluating concordance based on acquired mutations, copy number alterations and clonal structure.

Results: Between November 2016 - October 2021, 379 PDX models were developed (265 distinct models) representing 69 discrete diagnoses. Sarcoma represents the most common model type (50 discrete osteosarcoma, 20 desmoplastic small round cell tumor, 14 Ewing sarcoma, 24 rhabdomyosarcoma, 2 CIC/DUX4 and 2 BCOR-rearranged sarcoma) followed by neuroblastoma (n=35), leukemia (n=44), and Wilms tumor (n=15). While the majority of PDXs were established from recurrent or metastatic tissue, 7 paired diagnostic/pre-therapy and post-therapy or relapse models were generated. Genomic characterization of PDXs demonstrate excellent concordance and recapitulation of single nucleotide variants (90%), structural (88%) and copy number variants (94%) between patient tumor and matched PDX. Discrepancies between matched patient/PDX pairs are due to sub-clonal heterogeneity in source tumors with clonal outgrowth in the PDX. Analysis of serial PDX passages also demonstrate stable recapitulation of the genomic profile. Establishment of a diverse PDX collection allowed preclinical evaluation of 10 targeted agents across a spectrum of pediatric tumors and provided the preclinical rationale for 3 investigator-initiated pediatric clinical trials.

Conclusions: Investment in the development of a phenotypically diverse and biologically faithful collection of pediatric PDX models enables the goals of precision medicine. Optimization of PDX workflows and methods has also enabled the development of a pediatric PDX consortium (PROXC - Pediatric Research in Oncology Xenografting Consortium) to further support the development of pre- and post-clinical studies for pediatric cancer.

Citation Format: Filemon S. Dela Cruz, Joseph G. McCarter, Daoqi You, Nancy Bouvier, Xinyi Wang, Kristina C. Guillan, Armaan H. Siddiquee, Katie B. Souto, Hongyan Li, Teng Gao, Dominik Glodzik, Daniel Diolaiti, Neerav N. Shukla, Joachim Silber, Umeshkumar K. Bhanot, Faruk Erdem Kombak, Diego F. Coutinho, Shanita Li, Juan E. Arango Ossa, Juan S. Medina-Martinez, Michael V. Ortiz, Emily K. Slotkin, Michael D. Kinnaman, Sameer F. Sait, Tara J. O'Donohue, Marissa Mattar, Maximiliano Meneses, Michael P. LaQuaglia, Todd E. Heaton, Justin T. Gerstle, Nicola Fabbri, Chelsey M. Burke, Irene M. Rodriquez-Sanchez, Christine A. Iacobuzio-Donahue, Julia L. Glade Bender, Ryan D. Roberts, Jason T. Yustein, Nino C. Rainusso, Brian D. Crompton, Elizabeth Stewart, Alejandro Sweet-Cordero, Leanne C. Sayles, Andrika D. Thomas, Michael H. Roehrl, Elisa de Stanchina, Elli Papaemmanuil, Andrew L. Kung. Development of a patient-derived xenograft (PDX) modeling program to enable pediatric precision medicine [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 704.

Subclonal somatic copy number alterations emerge and dominate in recurrent osteosarcoma

11533

Background: Multiple large-scale tumor genomic profiling efforts have been undertaken in osteosarcoma, however little is known about the spatial and temporal intratumor heterogeneity and how it may drive treatment resistance. Methods: We performed 30-80x whole genome sequencing (WGS) of 37 tumor samples from 8 patients with relapsed or refractory osteosarcoma. A set of high confidence single nucleotide variants (SNV), copy number alterations (CNA), structural variations (SV) were called for each sample using our pediatric expanded genomics pipeline and an evolutionary analysis was performed using a custom pipeline of computational tools. Results: Of the 8 patients in our cohort, 4 had localized disease at diagnosis (OSCE4, OSCE5, OSCE6, OSCE9) and 4 had metastatic disease at diagnosis (OSCE1, OSCE2, OSCE3, OSCE10). There were 17 samples from primary sites, 7 were pretreatment biopsies, 10 from on therapy primary resections. 20 samples came from metastatic sites, 15 of which were from lung metastases. Driver gene SNV’s were identified in 5 of 8 patients, including TP53 (OSCE1), ATRX (OSCE3, OSCE10), RB1 (OSCE4), and CDKN2A (OSCE9). No new driver SNV’s emerged post-therapy in any patient. HATCHet, an algorithm which infers clone specific copy number alterations, identified subclonal CNAs in all but one patient (OSCE2). In the 7 patients with subclonal CNAs, 6 had two copy number clones identified, and 1 patient (OSCE10) had three copy number clones identified. In 5 patients (OSCE1, OSCE4, OSCE5, OSCE6, OSCE10) there is a copy number clone that is subclonal in the primary tumor which emerges and dominates at subsequent relapses. The resistant clone in each of these cases had either MYC gain/amplification. Amplifications in CCNE1 (OSCE1), RAD21 (OSCE4, OSCE5, OSCE10), VEGFA (OSCE1, OSCE9), IGF1R (OSCE6) were also identified as potential drivers in the resistant copy number clones. In two of these patients (OSCE1, OSCE6), the treatment resistant subclone becomes the dominant copy number clone by the time of primary resection. SNV based phylogenies revealed monoclonal and polyclonal seeding of metastases and monophyletic and polyphyletic modes of dissemination. Over half the new mutations acquired in recurrent disease were attributed to HRD or cisplatin mutational signatures. Conclusions: Subclonal copy number clones emerge and dominate in relapsed osteosarcoma, with MYC gain/amplification a defining characteristic in our cohort. Selective pressure from neoadjuvant chemotherapy reveals this clone at the time of primary resection, implying genomic profiling at this timepoint may be more reflective of its metastatic potential. [Table: see text]

Immunogenomic characterization of biliary tract cancers: Biomarker enrichment for benefit to immune checkpoint blockade

4083

Background: Several immunomodulatory molecules (PD-L1, B7H4, and CD276) have been associated with biliary tract cancer (BTC) subgroups, suggesting potential value to immune checkpoint blockade (ICB) in this lethal disease. Phase II monotherapy (pembrolizumab or nivolumab), and combination (atezolizumab and cobimetinib) ICB trials reported low response rates in unselected advanced BTC with a wide range of responses. A recent randomized phase III trial (TOPAZ-1) reported an overall survival (OS) benefit among patients (pts) with advanced BTC treated with chemotherapy and anti-PD-L1 ICB. However, no correlation between PD-L1 expression and OS was noted and biomarker enrichment strategy in BTC for immunotherapy remains a key to optimize OS. Methods: From our comprehensive clinico-genomic database for BTC at Memorial Sloan Kettering (MSK), a retrospective genomic landscape and neoantigen analysis was performed using MSK-IMPACT. Potential immunogenic subgroups were evaluated: homologous recombination deficiency (HRD) defined by pathogenic alterations in BRCA1/2, PALB2, and BAP1, microsatellite stability high (MSI-H) defined by MSIsensor score ≥10, and tumor mutation burden (TMB)>10. Clinical outcomes with anti-PD-1 ICB were evaluated. Results: Among N=1,190 pts with BTC, N=1,346 samples were sequenced between 03/2014 and 01/2022. Key actionable alterations included (%): IDH1, 2 (13, 3), FGFR2 fusions (9), ERBB2 amplification (5), BRAF V600E (2), RNF43 (2), POLE (2), NTRK1 fusion (<1). There were N=230 (17%) patients with putatively more immunogenic BTC (iBTC) identified by HRD [ BRCA1/2 (1, 2.4), PALB2 (1), BAP1 (9)], TMB>10, and MSI-H. Frequency, location (intrahepatic, ICC; extrahepatic, ECC; gallbladder, GBC), TMB, and genomic instability score (GIS) are summarized (Table). Among iBTC subgroup, N=32 pts received ICB. Their median follow up was 29.1 months. Median lines of prior therapy was 3. Median PFS was 5.6 M (95%CI: 1.2-10.1) and OS was 33.4 M (23.1-43.6). Conclusion: A subgroup of BTC pts (iBTC) benefit from ICB. Apart from MSI-H and TMB>10, other genomically-defined subgroups such as HRD may benefit from ICB. Prospective studies are needed to evaluate a better biomarker enrichment strategy beyond PD-L1 and TMB, that can represent other immunogenic aspects of tumor neoantigen and microenvironment. [Table: see text]

Evidence for reduced BRCA2 functional activity in Homo sapiens after divergence from the chimpanzee-human last common ancestor

We performed a comparative analysis of human and 12 non-human primates to identify sequence variations in known cancer genes. We identified 395 human-specific fixed non-silent substitutions that emerged during evolution of human. Using bioinformatics analyses for functional consequences, we identified a number of substitutions that are predicted to alter protein function; one of these mutations is located at the most evolutionarily conserved domain of human BRCA2.

Genomic and transcriptomic analysis of a library of small cell lung cancer patient-derived xenografts

Access to clinically relevant small cell lung cancer (SCLC) tissue is limited because surgical resection is rare in metastatic SCLC. Patient-derived xenografts (PDX) and circulating tumor cell-derived xenografts (CDX) have emerged as valuable tools to characterize SCLC. Here, we present a resource of 46 extensively annotated PDX/CDX models derived from 33 patients with SCLC. We perform multi-omic analyses, using targeted tumor next-generation sequencing, RNA-sequencing, and immunohistochemistry to deconvolute the mutational landscapes, global expression profiles, and molecular subtypes of these SCLC models. SCLC subtypes characterized by transcriptional regulators, ASCL1, NEUROD1 and POU2F3 are confirmed in this cohort. A subset of SCLC clinical specimens, including matched PDX/CDX and clinical specimen pairs, confirm that the primary features and genomic and proteomic landscapes of the tumors of origin are preserved in the derivative PDX models. This resource provides a powerful system to study SCLC biology.

Genomic characterization of metastatic patterns from prospective clinical sequencing of 25,000 patients

Metastatic progression is the main cause of death in cancer patients, whereas the underlying genomic mechanisms driving metastasis remain largely unknown. Here, we assembled MSK-MET, a pan-cancer cohort of over 25,000 patients with metastatic diseases. By analyzing genomic and clinical data from this cohort, we identified associations between genomic alterations and patterns of metastatic dissemination across 50 tumor types. We found that chromosomal instability is strongly correlated with metastatic burden in some tumor types, including prostate adenocarcinoma, lung adenocarcinoma, and HR+/HER2+ breast ductal carcinoma, but not in others, including colorectal cancer and high-grade serous ovarian cancer, where copy-number alteration patterns may be established early in tumor development. We also identified somatic alterations associated with metastatic burden and specific target organs. Our data offer a valuable resource for the investigation of the biological basis for metastatic spread and highlight the complex role of chromosomal instability in cancer progression.

MYC levels regulate metastatic heterogeneity in pancreatic adenocarcinoma

The degree of metastatic disease varies widely amongst cancer patients and impacts clinical outcomes. However, the biological and functional differences that drive the extent of metastasis are poorly understood. We analyzed primary tumors and paired metastases using a multi-fluorescent lineage-labeled mouse model of pancreatic ductal adenocarcinoma (PDAC) - a tumor type where most patients present with metastases. Genomic and transcriptomic analysis revealed an association between metastatic burden and gene amplification or transcriptional upregulation of MYC and its downstream targets. Functional experiments showed that MYC promotes metastasis by recruiting tumor associated macrophages (TAMs), leading to greater bloodstream intravasation. Consistent with these findings, metastatic progression in human PDAC was associated with activation of MYC signaling pathways and enrichment for MYC amplifications specifically in metastatic patients. Collectively, these results implicate MYC activity as a major determinant of metastatic burden in advanced PDAC.

Early-Onset Pancreas Cancer: Clinical Descriptors, Genomics, and Outcomes

BACKGROUND

Recent evidence suggests a rising incidence of cancer in younger individuals. Herein, we report the epidemiologic, pathologic, and molecular characteristics of a patient cohort with early-onset pancreas cancer (EOPC).

METHODS

Institutional databases were queried for demographics, treatment history, genomic results, and outcomes. Overall survival from date of diagnosis was estimated using Kaplan-Meier method.

RESULTS

Between 2008 and 2018, 450 patients with EOPC were identified at Memorial Sloan Kettering. Median overall survival was 16.3  (95% confidence interval [CI] = 14.6 to 17.7) months in the entire cohort and 11.3  (95% CI = 10.2 to 12.2) months for patients with stage IV disease at diagnosis. Of the patients, 132 (29.3% of the cohort) underwent somatic testing; 21 of 132 (15.9%) had RAS wild-type cancers with identification of several actionable alterations, including ETV6-NTRK3, TPR-NTRK1, SCLA5-NRG1, and ATP1B1-NRG1 fusions, IDH1 R132C mutation, and mismatch repair deficiency. A total of 138 patients (30.7% of the cohort) underwent germline testing; 44 of 138 (31.9%) had a pathogenic germline variant (PGV), and 27.5% harbored alterations in cancer susceptibility genes. Of patients seen between 2015 and 2018, 30 of 193 (15.5%) had a PGV. Among 138 who underwent germline testing, those with a PGV had a reduced all-cause mortality compared with patients without a PGV controlling for stage and year of diagnosis (hazard ratio = 0.42, 95% CI = 0.26 to 0.69).

CONCLUSIONS

PGVs are present in a substantial minority of patients with EOPC. Actionable somatic alterations were identified frequently in EOPC, enriched in the RAS wild-type subgroup. These observations underpin the recent guidelines for universal germline testing and somatic profiling in pancreatic ductal adenocarcinoma.

The mutational landscape of human somatic and germline cells

Over the course of an individual's lifetime, normal human cells accumulate mutations1. Here we compare the mutational landscape in 29 cell types from the soma and germline using multiple samples from the same individuals. Two ubiquitous mutational signatures, SBS1 and SBS5/40, accounted for the majority of acquired mutations in most cell types, but their absolute and relative contributions varied substantially. SBS18, which potentially reflects oxidative damage2, and several additional signatures attributed to exogenous and endogenous exposures contributed mutations to subsets of cell types. The rate of mutation was lowest in spermatogonia, the stem cells from which sperm are generated and from which most genetic variation in the human population is thought to originate. This was due to low rates of ubiquitous mutational processes and may be partially attributable to a low rate of cell division in basal spermatogonia. These results highlight similarities and differences in the maintenance of the germline and soma.

Pancreas cancer and BRCA: A critical subset of patients with improving therapeutic outcomes

BACKGROUND

Patients with germline/somatic BRCA1/BRCA2 mutations (g/sBRCA1/2) comprise a distinct biologic subgroup of pancreas ductal adenocarcinoma (PDAC).

METHODS

Institutional databases were queried to identify patients who had PDAC with g/sBRCA1/2. Demographics, clinicopathologic details, genomic data (annotation sBRCA1/2 according to a precision oncology knowledge base for somatic mutations), zygosity, and outcomes were abstracted. Overall survival (OS) was estimated using the Kaplan-Meier method.

RESULTS

In total, 136 patients with g/sBRCA1/2 were identified between January 2011 and June 2020. Germline BRCA1/2 (gBRCA1/2) mutation was identified in 116 patients (85%). Oncogenic somatic BRCA1/2 (sBRCA1/2) mutation was present in 20 patients (15%). Seventy-seven patients had biallelic BRCA1/2 mutations (83%), and 16 (17%) had heterozygous mutations. Sixty-five patients with stage IV disease received frontline platinum therapy, and 52 (80%) had a partial response. The median OS for entire cohort was 27.6 months (95% CI, 24.9-34.5 months), and the median OS for patients who had stage IV disease was 23 months (95% CI, 19-26 months). Seventy-one patients received a poly(adenosine diphosphate ribose) polymerase (PARP) inhibitor (PARPi), and 52 received PARPi monotherapy. For maintenance PARPi, 10 patients (36%) had a partial response, 12 (43%) had stable disease, and 6 (21%) had progression of disease as their best response. Six patients (21%) received maintenance PARPi for >2 years. For those with stage IV disease who received frontline platinum, the median OS was 26 months (95% CI, 20-52 months) for biallelic patients (n = 39) and 8.66 months (95% CI, 6.2 months to not reached) for heterozygous patients (n = 4). The median OS for those who received PARPi therapy was 26.5 months (95% CI, 24-53 months) for biallelic patients (n = 25) and 8.66 months (95% CI, 7.23 months to not reached) for heterozygous patients (n = 2).

CONCLUSIONS

g/sBRCA1/2 mutations did not appear to have different actionable utility. Platinum and PARPi therapies offer therapeutic benefit, and very durable outcomes are observed in a subset of patients who have g/sBRCA1/2 mutations with biallelic status.

Abstract 703: An inflammatory leptomeningeal signature in cancer patients with neurologic manifestations of COVID-19

Background: COVID-19 is associated with a wide spectrum of neurologic manifestations, which can emerge weeks to months after the initial infection. Cancer patients are at a heightened risk of severe infections due to their immunocompromised status. Leading hypotheses predict that SARS-CoV-2 neuroinvasion or neurologic toxicity from the systemic cytokine storm may account for neurologic dysfunction.

Methods: We prospectively evaluated cancer patients with confirmed SARS-CoV-2 infection and subsequent neurologic manifestations at an NCI-designated cancer center. Evaluations included neurologic examinations, brain imaging, electroencephalogram, and cerebrospinal fluid (CSF) analysis for SARS-CoV-2 detection by RT-PCR for viral RNA and ELISA for nucleocapsid (N) and spike (S2) proteins. Additional proteomic analysis compared the CSF of patients with COVID-19 (CoV+) to cancer- and brain metastasis-matched COVID-19-negative controls (CoV-), and to patients with other cancer-associated neuroinflammatory conditions, including immune effector cell-associated neurotoxicity syndrome (ICANS) and autoimmune encephalitis.

Results: Between May and July 2020, we evaluated 18 COVID-19-positive cancer patients with a wide range of solid tumor and hematologic malignancies. Thirteen (72.2%) of our patients received tumor-directed treatment within 30 days of SARS-CoV-2 infection. Neurologic diagnoses include protracted critical care delirium (N=10), limbic encephalitis (N=4), refractory headaches (N=2), rhombencephalitis (N=1), and large-territory infarctions (N=1). A median delay of 19 days (range 0-77) existed between onset of respiratory symptoms and neurologic manifestations. Among 13 patients who underwent CSF analysis, there was no evidence of SARS-CoV-2 neuroinvasion by RT-PCR in the CSF (N=13) or by N and S2 protein detection (N=10). Targeted proteomic analysis detected a significant accumulation of IL-6 and -8, IFN-gamma, CXCL-1, -6, -9, -10, and -11, CCL-8 and -20, MMP-10 and 4E-BP1 in the CSF of CoV+ patients (N=10) relative to matched CoV- controls (pooled, p=0.029). Combined analysis of these 12 inflammatory mediators revealed CoV+ cytokine levels approaching that of ICANS. CSF MMP-10, a marker of neurodegeneration, correlated with neurologic dysfunction by Karnofsky performance status (p=0.011) and by disability rating scale (p=0.086) at the time of lumbar puncture.

Conclusion: A durable accumulation of IFN-gamma-mediated cytokines is detected in the CSF of cancer patients with neurologic manifestations of COVID-19, in the absence of detectable neuroinvasion. We hypothesize that cytokine-mediated neuroinflammation is responsible for the prolonged neurologic sequela of COVID-19. Our findings suggest a potential role for anti-inflammatory treatments in the management of such patients.

Citation Format: Jessica A. Wilcox, Jan Remsik, N. Esther Babady, Tracy A. McMillen, Behroze A. Vachha, Neil A. Halpern, Vikram Dhawan, Marc Rosenblum, Christine A. Iacobuzio-Donahue, Edward K. Avila, Bianca Santomasso, Adrienne Boire. An inflammatory leptomeningeal signature in cancer patients with neurologic manifestations of COVID-19 [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 703.

Multi-omic analysis of lung tumors defines pathways activated in neuroendocrine transformation

Lineage plasticity is implicated in treatment resistance in multiple cancers. In lung adenocarcinomas (LUADs) amenable to targeted therapy, transformation to small cell lung cancer (SCLC) is a recognized resistance mechanism. Defining molecular mechanisms of neuroendocrine (NE) transformation in lung cancer has been limited by a paucity of pre-/post-transformation clinical samples. Detailed genomic, epigenomic, transcriptomic, and protein characterization of combined LUAD/SCLC tumors, as well as pre-/post-transformation samples, support that NE transformation is primarily driven by transcriptional reprogramming rather than mutational events. We identify genomic contexts in which NE transformation is favored, including frequent loss of the 3p chromosome arm. We observed enhanced expression of genes involved in PRC2 complex and PI3K/AKT and NOTCH pathways. Pharmacological inhibition of the PI3K/AKT pathway delayed tumor growth and NE transformation in an EGFR-mutant patient-derived xenograft model. Our findings define a novel landscape of potential drivers and therapeutic vulnerabilities of neuroendocrine transformation in lung cancer.

Pancreatic cancer stem cells may define tumor stroma characteristics and recurrence patterns in pancreatic ductal adenocarcinoma

BACKGROUND

Herein, we investigate the relationship between pancreatic stem cell markers (PCSC markers), CD44, and epithelial-specific antigen (ESA), tumor stroma, and the impact on recurrence outcomes in pancreatic ductal adenocarcinoma (PDAC) patients.

METHODS

PDAC patients who underwent surgical resection between 01/2012-06/2014 were identified. CD44 and ESA expression was assessed by immunohistochemistry. Stroma was classified as loose, moderate, and dense based on fibroblast content. Overall survival (OS) and relapse-free survival (RFS) were estimated using the Kaplan-Meier method and compared between subgroups by log-rank test. The association between PCSC markers and stroma type was assessed by Fisher's exact test.

RESULTS

N = 93 PDAC patients were identified. The number of PDAC patients with dense, moderate density, and loose stroma was 11 (12%), 51 (54%), and 31 (33%) respectively. PDAC with CD44+/ESA- had highest rate of loose stroma (63%) followed by PDAC CD44+/ESA+ (50%), PDAC CD44-/ESA+ (35%), CD44-/ESA- (9%) (p = 0.0033). Conversely, lack of CD44 and ESA expression was associated with the highest rate of moderate and dense stroma (91% p = 0.0033). No local recurrence was observed in patients with dense stroma and 9 had distant recurrence. The highest rate of cumulative local recurrence was observed in patients with loose stroma. No statistically significant difference in RFS and OS was observed among subgroups (P = 0.089).

CONCLUSIONS

These data indicate PCSCs may have an important role in stroma differentiation in PDAC. Our results further suggest that tumor stroma may influence the recurrence pattern in PDAC patients.

The Genetic Evolution of Treatment-Resistant Cutaneous, Acral, and Uveal Melanomas

PURPOSE

Melanoma is a biologically heterogeneous disease composed of distinct clinicopathologic subtypes that frequently resist treatment. To explore the evolution of treatment resistance and metastasis, we used a combination of temporal and multilesional tumor sampling in conjunction with whole-exome sequencing of 110 tumors collected from 7 patients with cutaneous (n = 3), uveal (n = 2), and acral (n = 2) melanoma subtypes.

EXPERIMENTAL DESIGN

Primary tumors, metastases collected longitudinally, and autopsy tissues were interrogated. All but 1 patient died because of melanoma progression.

RESULTS

For each patient, we generated phylogenies and quantified the extent of genetic diversity among tumors, specifically among putative somatic alterations affecting therapeutic resistance.

CONCLUSIONS

In 4 patients who received immunotherapy, we found 1-3 putative acquired and intrinsic resistance mechanisms coexisting in the same patient, including mechanisms that were shared by all tumors within each patient, suggesting that future therapies directed at overcoming intrinsic resistance mechanisms may be broadly effective.

Germ cell tumors and associated hematologic malignancies evolve from a common shared precursor

Germ cell tumors (GCTs) are the most common cancer in men between the ages of 15 and 40. Although most patients are cured, those with disease arising in the mediastinum have distinctly poor outcomes. One in every 17 patients with primary mediastinal nonseminomatous GCTs develop an incurable hematologic malignancy and prior data intriguingly suggest a clonal relationship exists between hematologic malignancies and GCTs in these cases. To date, however, the precise clonal relationship between GCTs and the diverse additional somatic malignancies arising in such individuals have not been determined. Here, we traced the clonal evolution and characterized the genetic features of each neoplasm from a cohort of 15 patients with GCTs and associated hematologic malignancies. We discovered that GCTs and hematologic malignancies developing in such individuals evolved from a common shared precursor, nearly all of which harbored allelically imbalanced p53 and/or RAS pathway mutations. Hematologic malignancies arising in this setting genetically resembled mediastinal GCTs rather than de novo myeloid neoplasms. Our findings argue that this scenario represents a unique clinical syndrome, distinct from de novo GCTs or hematologic malignancies, initiated by an ancestral precursor that gives rise to the parallel evolution of GCTs and blood cancers in these patients.

Inflammatory Leptomeningeal Cytokines Mediate COVID-19 Neurologic Symptoms in Cancer Patients

SARS-CoV-2 infection induces a wide spectrum of neurologic dysfunction that emerges weeks after the acute respiratory infection. To better understand this pathology, we prospectively analyzed of a cohort of cancer patients with neurologic manifestations of COVID-19, including a targeted proteomics analysis of the cerebrospinal fluid. We find that cancer patients with neurologic sequelae of COVID-19 harbor leptomeningeal inflammatory cytokines in the absence of viral neuroinvasion. The majority of these inflammatory mediators are driven by type II interferon and are known to induce neuronal injury in other disease states. In these patients, levels of matrix metalloproteinase-10 within the spinal fluid correlate with the degree of neurologic dysfunction. Furthermore, this neuroinflammatory process persists weeks after convalescence from acute respiratory infection. These prolonged neurologic sequelae following systemic cytokine release syndrome lead to long-term neurocognitive dysfunction. Our findings suggest a role for anti-inflammatory treatment(s) in the management of neurologic complications of COVID-19 infection.

Initial Whole-Genome Sequencing of Plasma Cell Neoplasms in First Responders and Recovery Workers Exposed to the World Trade Center Attack of September 11, 2001

PURPOSE

The World Trade Center (WTC) attack of September 11, 2001 created an unprecedented environmental exposure to known and suspected carcinogens. High incidence of multiple myeloma and precursor conditions has been reported among first responders to the WTC disaster. To expand on our prior screening studies, and to characterize the genomic impact of the exposure to known and potential carcinogens in the WTC debris, we were motivated to perform whole-genome sequencing (WGS) of WTC first responders and recovery workers who developed a plasma cell disorder after the attack.

EXPERIMENTAL DESIGN

We performed WGS of nine CD138-positive bone marrow mononuclear samples from patients who were diagnosed with plasma cell disorders after the WTC disaster.

RESULTS

No significant differences were observed in comparing the post-WTC driver and mutational signature landscapes with 110 previously published WGSs from 56 patients with multiple myeloma and the CoMMpass WGS cohort (n = 752). Leveraging constant activity of the single-base substitution mutational signatures 1 and 5 over time, we estimated that tumor-initiating chromosomal gains were windowed to both pre- and post-WTC exposure.

CONCLUSIONS

Although limitations in sample size preclude any definitive conclusions, our findings suggest that the observed increased incidence of plasma cell neoplasms in this population is due to complex and heterogeneous effects of the WTC exposure that may have initiated or contributed to progression of malignancy.

Unbiased in vivo preclinical evaluation of anticancer drugs identifies effective therapy for the treatment of pancreatic adenocarcinoma

Pancreatic ductal adenocarcinoma (PDAC) is typically diagnosed at an advanced stage, which limits surgical options and portends a dismal prognosis. Current oncologic PDAC therapies confer marginal benefit and, thus, a significant unmet clinical need exists for new therapeutic strategies. To identify effective PDAC therapies, we leveraged a syngeneic orthotopic PDAC transplant mouse model to perform a large-scale, in vivo screen of 16 single-agent and 41 two-drug targeted therapy combinations in mice. Among 57 drug conditions screened, combined inhibition of heat shock protein (Hsp)-90 and MEK was found to produce robust suppression of tumor growth, leading to an 80% increase in the survival of PDAC-bearing mice with no significant toxicity. Mechanistically, we observed that single-agent MEK inhibition led to compensatory activation of resistance pathways, including components of the PI3K/AKT/mTOR signaling axis, which was overcome with the addition of HSP90 inhibition. The combination of HSP90(i) + MEK(i) was also active in vitro in established human PDAC cell lines and in vivo in patient-derived organoid PDAC transplant models. These findings encourage the clinical development of HSP90(i) + MEK(i) combination therapy and highlight the power of clinically relevant in vivo model systems for identifying cancer therapies.

Inflammatory leptomeningeal cytokines mediate delayed COVID-19 encephalopathy

SARS-CoV-2 infection induces a wide spectrum of neurologic dysfunction. Here we show that a particularly vulnerable population with neurologic manifestations of COVID-19 harbor an influx of inflammatory cytokines within the cerebrospinal fluid in the absence of viral neuro-invasion. The majority of these inflammatory mediators are driven by type 2 interferon and are known to induce neuronal injury in other disease models. Levels of matrix metalloproteinase-10 within the spinal fluid correlate with the degree of neurologic dysfunction. Furthermore, this neuroinflammatory process persists weeks following convalescence from the acute respiratory infection. These prolonged neurologic sequelae following a systemic cytokine release syndrome lead to long-term neurocognitive dysfunction with a wide range of phenotypes.