Analytical validation of NeXT Personal®, an ultra-sensitive personalized circulating tumor DNA assay

We describe the analytical validation of NeXT Personal®, an ultra-sensitive, tumor-informed circulating tumor DNA (ctDNA) assay for detecting residual disease, monitoring therapy response, and detecting recurrence in patients diagnosed with solid tumor cancers. NeXT Personal uses whole genome sequencing of tumor and matched normal samples combined with advanced analytics to accurately identify up to ~1,800 somatic variants specific to the patient's tumor. A personalized panel is created, targeting these variants and then used to sequence cell-free DNA extracted from patient plasma samples for ultra-sensitive detection of ctDNA. The NeXT Personal analytical validation is based on panels designed from tumor and matched normal samples from two cell lines, and from 123 patients across nine cancer types. Analytical measurements demonstrated a detection threshold of 1.67 parts per million (PPM) with a limit of detection at 95% (LOD95) of 3.45 PPM. NeXT Personal showed linearity over a range of 0.8 to 300,000 PPM (Pearson correlation coefficient = 0.9998). Precision varied from a coefficient of variation of 12.8% to 3.6% over a range of 25 to 25,000 PPM. The assay targets 99.9% specificity, with this validation study measuring 100% specificity and in silico methods giving us a confidence interval of 99.92 to 100%. In summary, this study demonstrates NeXT Personal as an ultra-sensitive, highly quantitative and robust ctDNA assay that can be used to detect residual disease, monitor treatment response, and detect recurrence in patients.

Abstract 6668: Immune infiltrate co-occurrence and neoantigen similarity are prognostic factors in early stage NSCLC

The prevalence of early-stage non-small cell lung cancer (NSCLC) with curative treatment options is expected to increase with recent implementation of annual screening programs. Predictors and molecular drivers of disease relapse, especially the role of intra-tumoral immune dysfunction, remains unclear but critical for the refinement of therapeutic decisions. By leveraging a comprehensive individual portrait of each patient's immune system potential novel mechanisms associated with tumor relapse in early-stage NSCLC may be identified. We profiled 11 non-relapsed (at least 2 year FU) lung adenocarcinoma patients and 11 covariate-matched (gender, age, stage) relapsed patients, who underwent curative treatment in stage IA-IIIB disease. We used NeXT SummitTM for variant and CNA calling, gene expression quantification, neoantigen prediction, HLA profiling (typing, mutation, and loss of heterozygosity), T-cell receptor and tumor microenvironment (TME) profiling. Neoantigen peptide sequences were subjected to further filtering and clustering based on between-patient similarity scores, with the goal of identifying shared clusters of relapse-associated neoantigens in each possible pair of patients. Differential network analyses were applied to the TME composition estimates to investigate relapse-associated patterns of cellular co-occurrence and interaction. When considering neoantigens selected on the basis of similarity, we found that those belonging to non-relapsed patients had significantly lower HLA binding rank (17.8 points) compared to that of relapsed patients (P=0.02), indicating weaker binding for relapsed cases. Clustering of both the most similar and frequently shared neoantigens correlated with relapse (P < 0.002). In the TME, we observed differential immune cell co-occurrence associated with relapse status, such as Tregs are positively correlated with B and CD4 T cells only in relapsed patients (Pearson’s R=0.7 and 0.74, both P<0.02 vs. R=0.18 and 0.35, both P>0.2 in non-relapsed patients), indicating suppressive anti-tumor immunity. Relapsed patients did not share significant enrichment of mutations in any biological pathway. Surprisingly, mutation purity (less mutations than expected by chance) was observed in relapsed patients, suggesting selective killing and escape. In this pilot cohort, we used an integrated platform to broadly characterize both the tumor and immune system, enabling identification of relapse-associated neoantigens that may share universal features which enhance HLA binding. Relapses in early-stage LUAD patients were associated with neoantigens with lower immunogenicity and an immunosuppressive TME. These findings demonstrate that deeper profiling of shared neoantigen features has the potential to become an early biomarker of relapse, informing patient therapy selection and surveillance.

Citation Format: Martina M. Sykora, Jason Pugh, Bailiang Li, Finn O. Mildner, Hubert Hackl, Arno Amann, Fabienne I. Nocera, Rachel M. Pyke, Lee McDaniel, Charles W. Abbott, Sean M. Boyle, Richard O. Chen, Dominik Wolf, Sieghart Sopper, Gabriele Gamerith. Immune infiltrate co-occurrence and neoantigen similarity are prognostic factors in early stage NSCLC [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 6668.

Abstract 5472: Utilizing response in immune checkpoint inhibitor treated cohorts improves clinical applicability of neoantigen immunogenicity predictions

Introduction: Neoantigen-based biomarkers have improved predictions of response to immune checkpoint blockade (ICB) therapy, highlighting the importance of accurate prediction of immunogenic neoantigen candidates. A challenge in developing robust immunogenicity prediction models is limited availability of sequencing-associated immunogenicity data for evaluating methods due to the complexity of generating such datasets. We propose a novel approach to optimize prediction models of immunogenic neoantigens using a meta-analysis framework based on multiple ICB cohorts.

Methods: To build on the 110 mono-allelic immunopeptidomics-derived SHERPA® MHC binding prediction framework, we engineered T-cell recognition features on two datasets: peptide-centric data aggregated by Schmidt et al. and patient-specific exome and transcriptome sequencing data from the TESLA consortium. We developed two-tiered models based on the feature landscapes of both datasets to predict peptide-MHC (pMHC) immunogenicity, incorporating features with significant performance gains. We systematically re-processed publicly available DNA and RNA sequencing data from over 500 ICB treated patients spanning 12 different cohorts across five different cancer types with a harmonized bioinformatics pipeline. We then evaluated the performance of each model consisting of a unique combination of immunogenic features across the ICB training (N=7) and validation (N=5) cohorts using a meta-analysis framework.

Results: We evaluated iterations of SHERPA-Immunogenicity (SI) models using the Schmidt et al. and TESLA datasets, resulting in a range of performance metrics (area under the precision recall curves of 0.74-0.84 and positive predictive values of 0.32-0.54). After aggregating pMHC predictions into patient-specific scores based on the most immunogenic peptide present (SHERPA-Immunogenicity Maximum - SIM) or the quantity of immunogenic peptides identified (SHERPA-Immunogenicity Burden - SIB), we observed that responders had higher SIM and SIB scores compared to non-responders across the melanoma training cohorts. We found SIM scores outperformed SIB scores, suggesting the degree of epitope immunogenicity may be a critical factor in predicting response. The model with the most significant meta p-value for ICB response in melanoma cohorts (OR=2.43, p=0.006) also predicted overall survival in 3/5 melanoma cohorts (p&lt0.05).

Conclusions: We developed a novel framework to predict neoantigen immunogenicity utilizing meta-analysis of ICB cohorts to overcome dataset limitations and gain prediction performance confidence. We look forward to supporting personalized cancer vaccine development with our pMHC immunogenicity predictions and applying our predictive biomarker on additional ICB cohorts.

Citation Format: Hima Anbunathan, Neeraja Ravi, Rachel Marty Pyke, Steven Dea, Richard O. Chen, Sean Michael Boyle. Utilizing response in immune checkpoint inhibitor treated cohorts improves clinical applicability of neoantigen immunogenicity predictions. [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 5472.

Abstract 5021: Accurate quantification of infiltrating B cell composition and clone diversity in tumor samples

Tumors harbor a complex ecosystem of malignant, immune, and stromal cells. While malignant cells dictate much of the tumor biology, there is evidence that the tumor microenvironment (TME) also plays a major role in disease etiology. Given the complexity and abundance of the TME cellular composition, investigating the role of immune cell types will yield novel biomarkers for tumor progression and response to therapies.

The role of B cells as a prognostic biomarker remains elusive. For instance, infiltrating B cells in CRC have both positive and negative prognostic value. Thus, a scalable approach to quantify B cells and the B-cell receptor (BCR) repertoire could yield novel insights into the role of B cells in tumor biology. To address this, we have developed immune cell quantification (InfiltrateID࣪) and immune receptor repertoire profiling (RepertoireID࣪) methods as part of the ImmunoID NeXT Platform®, an augmented, immuno-oncology-optimized exome/transcriptome platform.

We estimate B cell abundance and BCR repertoire by profiling FFPE and PBMC samples using ImmunoID NeXT࣪. In expanding upon InfiltrateID to further estimate B cell abundance, here we regress the bulk RNA-seq readout from a reference signature from purified immune cell types. We also generate orthogonal quantifications of B cell abundance by profiling samples with cytometry by time of flight, single-cell RNA-seq, flow cytometry, and immunohistochemistry (IHC). We compare BCR results from ImmunoID NeXT to a standalone sequencing approach to evaluate the concordance of top clones. We then utilize BCR profiling from ImmunoID NeXT to analyze clonality and isotype composition in tumor samples.

We first use InfiltrateID to estimate absolute B cell fractions in over 50 samples. Overall, we observe a high correlation between InfiltrateID results and orthogonal data sets in both PBMC and tumor FFPE samples (R2=0.90). When comparing BCR results from RepertoireID to a standalone BCR sequencing method that profiles IgM and IgG, we identify 475 and 387 of the top 500 clones in IgG and IgM, respectively, with highly concordant abundances across all clones (R2>0.72 and R2>0.82 in IgM and IgG, respectively). Next, we use InfiltrateID to estimate absolute B cell fractions in over 650 samples from 14 tumor types. On average, samples display B cell fractions in agreement with the literature and IHC quantifications, with higher B cell fractions in lung, breast, and cervical tumors. We also observe a range of BCR clonality values across tumor types. Finally, we observe differences in B cell composition and repertoire diversity in tumor samples from patients who underwent checkpoint blockade therapy.

We show that InfiltrateID and RepertoireID accurately capture the composition and clone diversity of infiltrating B cells in tumor samples.

Citation Format: Fabio Navarro, Eric Levy, Pamela Milani, Qiang Li, Shruti Bhide, Upasana Dutta, Charles W. Abbott, Jose Jacob, Rena McClory, John West, John Lyle, Sean Boyle, Richard O. Chen. Accurate quantification of infiltrating B cell composition and clone diversity in tumor samples [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 5021.

Abstract 5640: Mono-allelic immunopeptidomics data from 109 MHC-I alleles reveals variability in binding preferences and improves neoantigen prediction algorithm

Neoantigen-based biomarkers are a promising approach for stratifying patient response to immunotherapy; however, current neoantigen prediction methods are not accurate enough to optimize these biomarkers. Sequence variability in the major histocompatibility complex (MHC) leads to the presentation of diverse neoantigens to T cells, and accurately representing this diversity in neoantigen prediction is critical for improvement. Previously, we published data from 25 mono-allelic cell lines and built an associated MHC class I, pan-allelic neoantigen prediction algorithm (SHERPATM). Here, we profile an additional 84 MHC alleles including 37 that have never previously been profiled with mono-allelic immunopeptidomics, explore the impact of MHC variability on peptide binding and improve neoantigen prediction of the SHERPA algorithm. To generate the data, we stably and transiently transfected 109 different MHC alleles (43 HLA-A, 56 -B and 10 -C alleles) into independent K562 HLA-null cell lines, immunoprecipitated intact MHC complexes using a W6/32 antibody and profiled the bound peptides using LC/MS-MS. We recovered a median of 1430 peptides per allele, with yields from the transient transfections being consistently higher than the stable transfections. Nearly all alleles have a strong anchor residue in the ninth position, but the positions of the secondary anchor residue vary by gene. HLA-B showed a stronger preference for the second position while HLA-A exhibited more variability across the first, second and third positions. In addition to the 109 mono-allelic cell lines, SHERPA increases generalizability by systematically integrating an additional 104 mono-allelic and 384 multi-allelic samples with publicly available immunopeptidomics data. The 186 alleles in the resulting training dataset have an average allelic coverage of 98% across 18 different US-based ethnicities. We evaluated our updated performance on 10% held-out mono-allelic test data from multiple cell line sources. The positive predictive value (PPV) of SHERPA was markedly higher than either NetMHCPan 4.1 or MHCFlurry-2.0 (1.45 and 1.58-fold increase, respectively), with further gains when only the 37 previously unprofiled alleles were considered (1.51 and 1.79-fold increase, respectively). Furthermore, the SHERPA model was able to detect 1.38-fold more immunogenic epitopes than either other method. Finally, we performed predictions with SHERPA across millions of synthetic binding pockets and peptides to elucidate the impact of MHC variability on peptide diversity. We found a strong correlation between binding pocket positions that highly influence peptide binding and those that are evolutionarily divergent. In conclusion, we profiled 109 mono-allelic cell lines, showed key trends in MHC-associated peptides and improved the SHERPA neoantigen prediction model.

Citation Format: Rachel Marty Pyke, Steven Dea, Hima Anbunathan, Charles W. Abbott, Neeraja Ravi, Jason Harris, Gabor Bartha, Sejal Desai, Rena McClory, John West, Michael P. Snyder, Richard O. Chen, Sean Michael Boyle. Mono-allelic immunopeptidomics data from 109 MHC-I alleles reveals variability in binding preferences and improves neoantigen prediction algorithm [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 5640.

Abstract 6385: Applying NeXT Liquid Biopsy™, an exome-scale platform, to monitor and discover somatic variants in a broad set of cancer types

Circulating tumor cell-free DNA (ctDNA) has become a biomarker for prognosis and disease monitoring. However, studies typically utilize assays limited to a small set of genes that may miss biologically important and clinically actionable mutations. To address this limitation, we have developed a whole-exome scale cfDNA platform, NeXT Liquid Biopsy (NeXT LB), that enables sensitive identification of mutations in plasma across ~20,000 genes following interventions such as surgery and treatment therapies. NeXT LB monitors tumor variants and discovers novel mutations in the plasma through analysis of tumor, normal, and plasma samples from the same patient.

To enable sensitive detection across the exome in solid tumor and liquid biopsies, we developed an enhanced whole-exome assay and chemistry that augments challenging genomic regions to enable more uniform coverage across the exome. Additionally, we achieve a mean depth of coverage of ~2,000X across the exome, with boosted depth (~5,000X) for 247 clinically relevant oncogenic or tumor suppressor genes to further enhance sensitivity. Finally, we developed computational algorithms to sensitively monitor and discover somatic mutations in liquid biopsies without compromising specificity.

In this work, we measure the sensitivity of NeXT LB using SeraCare reference samples with known variants at 0.5%, 1%, and 2% allele fraction (AF). We observe 100% sensitivity at 2% and 1% AF, and >95% sensitivity at 0.5% AF. Additionally, we measure >95% sensitivity for variants with AF >=2% using a proprietary cell-line media system.

We generate low-pass Whole Genome Sequencing (lpWGS) data to estimate ctDNA fraction in conjunction with NeXT LB. Considering tumor heterogeneity, NeXT LB is capable of monitoring and discovering somatic variants when lpWGS-reported ctDNA fraction is >=3%, thereby highlighting the performance of the NeXT LB platform.

We apply NeXT LB to sequence over 100 matched plasma and normal samples at 250 gigabases (G) and tumors at 50 G. This data demonstrate somatic variation in over 1,000 distinct genes across the cohort, thereby demonstrating the breadth and performance improvements provided by our exome-scale platform in contrast to existing targeted platforms. Additionally, we find that the plasma variants are enriched for higher AFs in solid tumors, thus allowing comprehensive coverage of driver genes and recapitulating hotspots identified in public datasets, including TCGA.

We developed an exome-scale NeXT LB technology that enables sensitive monitoring and detection of somatic SNVs and indels from cfDNA. The NeXT LB platform covers a much broader landscape of tumor mutations from the plasma than existing targeted platforms, thereby enabling more comprehensive monitoring and discovery of mutations related to therapies, mechanisms of resistance, intra- and inter-tumor heterogeneity, among others.

Citation Format: Fabio C p Navarro, Naveen Ramesh, Josette Northcott, Rui Chen, Lee D. McDaniel, Charles W. Abbott, Dan Norton, Robin Li, John Lyle, Jason Harris, Gabor Bartha, John West, Sean M. Boyle, Richard O. Chen. Applying NeXT Liquid Biopsy™, an exome-scale platform, to monitor and discover somatic variants in a broad set of cancer types [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 6385.

Abstract 5163: A high sensitivity, tumor-informed liquid biopsy platform, designed to detect minimal residual disease at part per million resolution

Tumor-informed liquid biopsy approaches have proven promising for detecting minimal residual disease (MRD) and recurrence of cancer following surgical resection or other therapy. However, current liquid biopsy MRD assays typically detect ctDNA in a range above 30 to 300 parts per million (PPM), leaving a significant fraction of MRD cases undetected, particularly soon after surgery and in early stage cancers where ctDNA can be at very low levels. To address this, we have developed NeXT Personal™, a tumor-informed liquid biopsy assay that achieves sensitivity down to 1 PPM, therefore enabling earlier detection of MRD and recurrence.

NeXT Personal leverages tumor/normal whole genome sequencing to design personalized MRD liquid biopsy panels for each patient. The panel is composed of >1,200 somatic tumor variants enabling higher sensitivity MRD detection in plasma through tracking of larger numbers of high quality and lower noise variants. This allows the platform to achieve high sensitivity across cancer types and stages, including early stage cancers and low mutational burden tumors, utilizing ~4 mL of plasma. Two independent methods were used to establish utility and performance: a proprietary cell-line media system, and well-characterized matched tumor-normal-plasma patient samples. Samples were serially diluted to <1 PPM, with replicates used to confirm performance. Digital droplet polymerase chain reaction (ddPCR) was used to orthogonally validate platform performance to the limit of detection (LOD) of ddPCR.

Characterization of MRD LOD in three cell-line media systems, HCC1143, HCC38, and HCC1937, yielded accurate and reproducible detection of signal across a broad range of concentrations, to a lower limit of 1-2 PPM. We then used our platform to characterize MRD LOD in a set of serially diluted patient samples, demonstrating sensitivity down to as low as 1 PPM, with high specificity in normal control samples. Finally, we demonstrated the performance of NeXT Personal with matched tumor-normal-plasma patient samples (8 different cancer types, stages II-IV). In this series, NeXT Personal detected cancers down to 0.8 PPM with high specificity demonstrated across a set of healthy normal donor samples. We estimate that ~50% of the cases in this set of patients would not have been detected by other commercially available liquid biopsy MRD platforms.

NeXT Personal achieved highly sensitive and specific MRD detection, reproducibly demonstrating a LOD down to 1 PPM in different cancer types and cell line dilutions, representing approximately 10 to 100 times higher sensitivity than other liquid biopsy MRD approaches. The high sensitivity of NeXT Personal potentially enables MRD detection across a broad variety of cancers and stages, including typically challenging early stage, low mutational burden, and low-shedding cancers.

Citation Format: Sean Michael Boyle, Gabor Bartha, John Lyle, Jason Harris, Josette Northcott, Dan Norton, Rachel Marty Pyke, Fabio C. P. Navarro, Alexander Stram, Christian Haudenschild, Rose Santiago, Robin Li, Chris Nelson, Yelia Huo, Manju Chinnappa, Qi Zhang, Lloyd Hsu, John West, Richard O. Chen. A high sensitivity, tumor-informed liquid biopsy platform, designed to detect minimal residual disease at part per million resolution [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 5163.

Abstract 5161: Exome-scale longitudinal tracking of emerging therapeutic resistance in GIST via analysis of circulating tumor DNA

Gastrointestinal stromal tumors (GIST) are lethal tumors characterized by constitutively activating mutations to KIT or PDGFRA. Transient disease control in the first-line setting is achieved via inhibition of tyrosine kinase signaling using the KIT inhibitor imatinib. As patients progress through subsequent lines of therapy a molecularly heterogeneous disease evolves, characterized by distinct subtypes and shifting repertoires of exon-specific KIT variants which directly impact treatment outcomes. Here, we use tumor-informed exome-scale liquid biopsy to identify and track the evolution of multiple resistance mechanisms in patients receiving tyrosine kinase inhibitors (TKIs) to address the unmet need of comprehensive understanding of GIST evolution in response to therapy.

Matched tumor, normal and serial plasma samples were obtained from 15 heavily pretreated metastatic GIST patients. Following baseline sample collection, all patients received systemic TKI therapy, and were monitored until disease progression. Exome-scale detection of somatic variants in cfDNA from longitudinal matched plasma samples was achieved using the NeXT Liquid BiopsyTM platform. The ImmunoID NeXT PlatformⓇ, an augmented exome/transcriptome platform and analysis pipeline which generates comprehensive tumor and immune data was used to profile paired tumor and normal samples.

Longitudinal whole exome sequencing of plasma identified dynamic shifts in existing clones harboring exon-specific KIT mutations, and evolution of new KIT mutations arising prior to identification of tumor progression using standard imaging techniques. We detected a correlation between the number of damaging mutations detected in baseline ctDNA and tumor exon 11 KIT mutation status, suggesting that plasma mutation profiles may be KIT-variant dependent. ctDNA from patients with shorter overall survival (OS) was enriched for variants in the PI3K-AKT and MAPK pathway, potentially contributing to immune evasion observed in those patients. Additional associations were observed between gene copy-number changes and OS (P = .0097). Previous studies have demonstrated that immune infiltration and activity may be KIT variant specific, here we broaden those findings, identifying a significant correlation between TCRɑ clonality and variants detected only in plasma (P = .04), as well as a significant association between TCRβ diversity and OS (HR = 2.55, log rank P = .04).

Comprehensive profiling of paired tumor tissue (WES and RNA-Seq) and WES of serially collected ctDNA sensitively and repeatedly identified evolving KIT mutations and other molecular alterations prior to radiologically confirmed disease progression. These findings suggest plasma-based monitoring of late-stage GIST malignancies may be useful for non-invasive disease tracking, providing treatment guidance prior to traditional approaches.

Citation Format: Charles W. Abbott, Niamh Coleman, Jing Wang, Josette Northcott, Jason Pugh, Dan Norton, Fábio C. Navarro, Lee D. McDaniel, Eric Levy, Rachel Marty Pyke, John Lyle, Jason Harris, Gabor Bartha, Filip Janku, John West, Richard O. Chen, Sean Boyle. Exome-scale longitudinal tracking of emerging therapeutic resistance in GIST via analysis of circulating tumor DNA [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 5161.

Sotigalimab and/or nivolumab with chemotherapy in first-line metastatic pancreatic cancer: clinical and immunologic analyses from the randomized phase 2 PRINCE trial

Chemotherapy combined with immunotherapy has improved the treatment of certain solid tumors, but effective regimens remain elusive for pancreatic ductal adenocarcinoma (PDAC). We conducted a randomized phase 2 trial evaluating the efficacy of nivolumab (nivo; anti-PD-1) and/or sotigalimab (sotiga; CD40 agonistic antibody) with gemcitabine/nab-paclitaxel (chemotherapy) in patients with first-line metastatic PDAC ( NCT03214250 ). In 105 patients analyzed for efficacy, the primary endpoint of 1-year overall survival (OS) was met for nivo/chemo (57.7%, P = 0.006 compared to historical 1-year OS of 35%, n = 34) but was not met for sotiga/chemo (48.1%, P = 0.062, n = 36) or sotiga/nivo/chemo (41.3%, P = 0.223, n = 35). Secondary endpoints were progression-free survival, objective response rate, disease control rate, duration of response and safety. Treatment-related adverse event rates were similar across arms. Multi-omic circulating and tumor biomarker analyses identified distinct immune signatures associated with survival for nivo/chemo and sotiga/chemo. Survival after nivo/chemo correlated with a less suppressive tumor microenvironment and higher numbers of activated, antigen-experienced circulating T cells at baseline. Survival after sotiga/chemo correlated with greater intratumoral CD4 T cell infiltration and circulating differentiated CD4 T cells and antigen-presenting cells. A patient subset benefitting from sotiga/nivo/chemo was not identified. Collectively, these analyses suggest potential treatment-specific correlates of efficacy and may enable biomarker-selected patient populations in subsequent PDAC chemoimmunotherapy trials.

A machine learning algorithm with subclonal sensitivity reveals widespread pan-cancer human leukocyte antigen loss of heterozygosity

Human leukocyte antigen loss of heterozygosity (HLA LOH) allows cancer cells to escape immune recognition by deleting HLA alleles, causing the suppressed presentation of tumor neoantigens. Despite its importance in immunotherapy response, few methods exist to detect HLA LOH, and their accuracy is not well understood. Here, we develop DASH (Deletion of Allele-Specific HLAs), a machine learning-based algorithm to detect HLA LOH from paired tumor-normal sequencing data. With cell line mixtures, we demonstrate increased sensitivity compared to previously published tools. Moreover, our patient-specific digital PCR validation approach provides a sensitive, robust orthogonal approach that could be used for clinical validation. Using DASH on 610 patients across 15 tumor types, we find that 18% of patients have HLA LOH. Moreover, we show inflated HLA LOH rates compared to genome-wide LOH and correlations between CD274 (encodes PD-L1) expression and microsatellite instability status, suggesting the HLA LOH is a key immune resistance strategy.

Human leukocyte antigen loss of heterozygosity (HLA LOH) is an important mechanism of immune escape in patients with cancer. Here the authors design and validate a machine learning algorithm with subclonal sensitivity for the identification of HLA LOH from paired tumor-normal sequencing data.

Ipilimumab alone or in combination with nivolumab in patients with advanced melanoma who have progressed or relapsed on PD-1 blockade: clinical outcomes and translational biomarker analyses

BACKGROUND

There are no validated biomarkers that can aid clinicians in selecting who would best benefit from anticytotoxic T lymphocyte-associated antigen 4 monotherapy versus combination checkpoint blockade in patients with advanced melanoma who have progressive disease after programmed death 1 (PD-1) blockade.

METHODS

We conducted a randomized multicenter phase II trial in patients with advanced melanoma. Patients were randomly assigned to receive either 1 mg/kg of nivolumab plus 3 mg/kg of ipilimumab or 3 mg/kg of ipilimumab every 3 weeks for up to four doses. Patients were stratified by histological subtype and prior response to PD-1 therapy. The primary clinical objective was overall response rate by week 18. Translational biomarker analyses were conducted in patients with blood and tissue samples.

RESULTS

Objective responses were seen in 5 of 9 patients in the ipilimumab arm and 2 of 10 patients in the ipilimumab+nivolumab arm; disease control rates (DCRs) (66.7% vs 60.0%) and rates of grade 3-4 adverse events (56% vs 50%) were comparable between arms. In a pooled analysis, patients with clinical benefit (CB), defined as Response Evaluation Criteria in Solid Tumors response or progression-free for 6 months, showed increased circulating CD4+ T cells with higher polyfunctionality and interferon gamma production following treatment. Tumor profiling revealed enrichment of NRAS mutations and activation of transcriptional programs associated with innate and adaptive immunity in patients with CB.

CONCLUSIONS

In patients with advanced melanoma that previously progressed on PD-1 blockade, objective responses were seen in both arms, with comparable DCRs. Findings from biomarker analyses provided hypothesis-generating signals for validation in future studies of larger patient cohorts.

TRIAL REGISTRATION NUMBER

NCT02731729.

CD40 agonistic monoclonal antibody APX005M (sotigalimab) and chemotherapy, with or without nivolumab, for the treatment of metastatic pancreatic adenocarcinoma: an open-label, multicentre, phase 1b study

BACKGROUND

Standard chemotherapy remains inadequate in metastatic pancreatic adenocarcinoma. Combining an agonistic CD40 monoclonal antibody with chemotherapy induces T-cell-dependent tumour regression in mice and improves survival. In this study, we aimed to evaluate the safety of combining APX005M (sotigalimab) with gemcitabine plus nab-paclitaxel, with and without nivolumab, in patients with pancreatic adenocarcinoma to establish the recommended phase 2 dose.

METHODS

This non-randomised, open-label, multicentre, four-cohort, phase 1b study was done at seven academic hospitals in the USA. Eligible patients were adults aged 18 years and older with untreated metastatic pancreatic adenocarcinoma, Eastern Cooperative Oncology Group performance status score of 0-1, and measurable disease by Response Evaluation Criteria in Solid Tumors version 1.1. All patients were treated with 1000 mg/m² intravenous gemcitabine and 125 mg/m² intravenous nab-paclitaxel. Patients received 0·1 mg/kg intravenous APX005M in cohorts B1 and C1 and 0·3 mg/kg in cohorts B2 and C2. In cohorts C1 and C2, patients also received 240 mg intravenous nivolumab. Primary endpoints comprised incidence of adverse events in all patients who received at least one dose of any study drug, incidence of dose-limiting toxicities (DLTs) in all patients who had a DLT or received at least two doses of gemcitabine plus nab-paclitaxel and one dose of APX005M during cycle 1, and establishing the recommended phase 2 dose of intravenous APX005M. Objective response rate in the DLT-evaluable population was a key secondary endpoint. This trial (PRINCE, PICI0002) is registered with ClinicalTrials.gov, NCT03214250 and is ongoing.

FINDINGS

Between Aug 22, 2017, and July 10, 2018, of 42 patients screened, 30 patients were enrolled and received at least one dose of any study drug; 24 were DLT-evaluable with median follow-up 17·8 months (IQR 16·0-19·4; cohort B1 22·0 months [21·4-22·7], cohort B2 18·2 months [17·0-18·9], cohort C1 17·9 months [14·3-19·7], cohort C2 15·9 months [12·7-16·1]). Two DLTs, both febrile neutropenia, were observed, occurring in one patient each for cohorts B2 (grade 3) and C1 (grade 4). The most common grade 3-4 treatment-related adverse events were lymphocyte count decreased (20 [67%]; five in B1, seven in B2, four in C1, four in C2), anaemia (11 [37%]; two in B1, four in B2, four in C1, one in C2), and neutrophil count decreased (nine [30%]; three in B1, three in B2, one in C1, two in C2). 14 (47%) of 30 patients (four each in B1, B2, C1; two in C2) had a treatment-related serious adverse event. The most common serious adverse event was pyrexia (six [20%] of 30; one in B2, three in C1, two in C2). There were two chemotherapy-related deaths due to adverse events: one sepsis in B1 and one septic shock in C1. The recommended phase 2 dose of APX005M was 0·3 mg/kg. Responses were observed in 14 (58%) of 24 DLT-evaluable patients (four each in B1, C1, C2; two in B2).

INTERPRETATION

APX005M and gemcitabine plus nab-paclitaxel, with or without nivolumab, is tolerable in metastatic pancreatic adenocarcinoma and shows clinical activity. If confirmed in later phase trials, this treatment regimen could replace chemotherapy-only standard of care in this population.

FUNDING

Parker Institute for Cancer Immunotherapy, Cancer Research Institute, and Bristol Myers Squibb.

Exome sequencing in 32 patients with anophthalmia/microphthalmia and developmental eye defects

Anophthalmia/microphthalmia (A/M) is a genetically heterogeneous birth defect for which the etiology is unknown in more than 50% of patients. We used exome sequencing with the ACE Exome(TM) (Personalis, Inc; 18 cases) and UCSF Genomics Core (21 cases) to sequence 28 patients with A/M and four patients with varied developmental eye defects. In the 28 patients with A/M, we identified de novo mutations in three patients (OTX2, p.(Gln91His), RARB, p.Arg387Cys and GDF6, p.Ala249Glu) and inherited mutations in STRA6 in two patients. In patients with developmental eye defects, a female with cataracts and cardiomyopathy had a de novo COL4A1 mutation, p.(Gly773Arg), expanding the phenotype associated with COL4A1 to include cardiomyopathy. A male with a chorioretinal defect, microcephaly, seizures and sensorineural deafness had two PNPT1 mutations, p.(Ala507Ser) and c.401-1G>A, and we describe eye defects associated with this gene for the first time. Exome sequencing was efficient for identifying mutations in pathogenic genes for which there is no clinical testing available and for identifying cases that expand phenotypic spectra, such as the PNPT1 and COL4A1-associated disorders described here.

The direct cellular target of topically applied pimecrolimus may not be infiltrating lymphocytes

BACKGROUND

Topically applied calcineurin inhibitors have been shown to be effective in the treatment of atopic dermatitis. When systemically administered, these agents cause immunosuppression via inhibition of calcineurin in lymphocytes. As topical agents, the mechanism of action is poorly defined.

OBJECTIVES

To test the hypothesis that skin-infiltrating lymphocytes are directly targeted when calcineurin inhibitors are applied to the skin.

METHODS

Ten patients with atopic dermatitis were treated with 1% pimecrolimus cream twice daily to target lesions. Skin biopsies were performed before and 48 h after beginning therapy. We assessed the cellular localization of NFAT1 and NFAT2 as a surrogate measure of intracellular calcineurin activity (e.g. increasing cytoplasmic localization with increasing calcineurin inhibition).

RESULTS

All patients showed a clinical response, at both 48 h and 2 weeks. As previously described, NFAT2 localized to the follicular keratinocytes, and its activation was partially inhibited by topical pimecrolimus. NFAT1 was found to be expressed by follicular and interfollicular keratinocytes, and its mostly nuclear localization was not affected by topical pimecrolimus therapy. Both NFAT1 and NFAT2 were found in the infiltrating lymphocytes. However, using both manual counting as well as an automated method to assess nuclear intensity of NFAT staining, we found that the proportion of infiltrating leucocytes with nuclear ('activated') NFAT did not change following therapy with pimecrolimus.

CONCLUSIONS

Our results suggest that topical pimecrolimus does not act primarily by inhibiting the calcineurin/NFAT axis in lymphocytes but may instead act by other mechanisms, possibly by decreasing NFAT2 activity in follicular keratinocytes.

A network-based analysis of systemic inflammation in humans

Oligonucleotide and complementary DNA microarrays are being used to subclassify histologically similar tumours, monitor disease progress, and individualize treatment regimens. However, extracting new biological insight from high-throughput genomic studies of human diseases is a challenge, limited by difficulties in recognizing and evaluating relevant biological processes from huge quantities of experimental data. Here we present a structured network knowledge-base approach to analyse genome-wide transcriptional responses in the context of known functional interrelationships among proteins, small molecules and phenotypes. This approach was used to analyse changes in blood leukocyte gene expression patterns in human subjects receiving an inflammatory stimulus (bacterial endotoxin). We explore the known genome-wide interaction network to identify significant functional modules perturbed in response to this stimulus. Our analysis reveals that the human blood leukocyte response to acute systemic inflammation includes the transient dysregulation of leukocyte bioenergetics and modulation of translational machinery. These findings provide insight into the regulation of global leukocyte activities as they relate to innate immune system tolerance and increased susceptibility to infection in humans.

Computational knowledge integration in biopharmaceutical research

An initiative to increase biopharmaceutical research productivity by capturing, sharing and computationally integrating proprietary scientific discoveries with public knowledge is described. This initiative involves both organisational process change and multiple interoperating software systems. The software components rely on mutually supporting integration techniques. These include a richly structured ontology, statistical analysis of experimental data against stored conclusions, natural language processing of public literature, secure document repositories with lightweight metadata, web services integration, enterprise web portals and relational databases. This approach has already begun to increase scientific productivity in our enterprise by creating an organisational memory (OM) of internal research findings, accessible on the web. Through bringing together these components it has also been possible to construct a very large and expanding repository of biological pathway information linked to this repository of findings which is extremely useful in analysis of DNA microarray data. This repository, in turn, enables our research paradigm to be shifted towards more comprehensive systems-based understandings of drug action.

Automated diagnosis of data-model conflicts using metadata

The authors describe a methodology for helping computational biologists diagnose discrepancies they encounter between experimental data and the predictions of scientific models. The authors call these discrepancies data-model conflicts. They have built a prototype system to help scientists resolve these conflicts in a more systematic, evidence-based manner. In computational biology, data-model conflicts are the result of complex computations in which data and models are transformed and evaluated. Increasingly, the data, models, and tools employed in these computations come from diverse and distributed resources, contributing to a widening gap between the scientist and the original context in which these resources were produced. This contextual rift can contribute to the misuse of scientific data or tools and amplifies the problem of diagnosing data-model conflicts. The authors' hypothesis is that systematic collection of metadata about a computational process can help bridge the contextual rift and provide information for supporting automated diagnosis of these conflicts. The methodology involves three major steps. First, the authors decompose the data-model evaluation process into abstract functional components. Next, they use this process decomposition to enumerate the possible causes of the data-model conflict and direct the acquisition of diagnostically relevant metadata. Finally, they use evidence statically and dynamically generated from the metadata collected to identify the most likely causes of the given conflict. They describe how these methods are implemented in a knowledge-based system called GRENDEL and show how GRENDEL can be used to help diagnose conflicts between experimental data and computationally built structural models of the 30S ribosomal subunit.

Standardized representations of the literature: combining diverse sources of ribosomal data

We are building a knowledge base (KB) of published structural data on the 30s ribosomal subunit in prokaryotes. Our KB is distinguished by a standardized representation of biological experiments and their results, in a reusable format. It can be accessed by computer programs that exploit the rich interconnections within the data. The KB is designed to support the construction of 3D models of the 30S subunit, as well as the analysis and extension of relevant functional and phylogenetic information. Most published information about the structure of the ubiquitous ribosome focuses on E. coli as a model system. At the same time, thousands of RNA sequences for the ribosome have been gathered and cataloged. The volume and complexity of these data can complicate attempts to separate structural data peculiar to E. coli from data of universal relevance. We have written an application that dynamically queries the KB and the Ribosome Database Project, a repository of ribosomal RNA sequences from other organisms, in order to assess the relevance of structural data to particular organisms. The application uses the RDP alignment to determine whether a set of data refer primarily to conserved, mismatched, or gapped positions. For a set of 16 representative articles evaluated over 211 sequences, 73% of observations have unambiguous translations from E. coli to the other organisms, 21% have somewhat ambiguous translations, and 6% have no translations. There is a wide variation in these numbers over different articles and organisms, confirming that some articles report structural information specific to E. coli while others report information that is quite general.

RIBOWEB: linking structural computations to a knowledge base of published experimental data

The world wide web (WWW) has become critical for storing and disseminating biological data. It offers an additional opportunity, however, to support distributed computation and sharing of results. Currently, computational analysis tools are often separated from the data in a manner that makes iterative hypothesis testing cumbersome. We hypothesize that the cycle of scientific reasoning (using data to build models, and evaluating models in light of data) can be facilitated with resources that link computations with semantic models of the data. Riboweb is an on-line knowledge-based resource that supports the creation of three-dimensional models of the 30S ribosomal subunit. It has three components: (I) a knowledge base containing representations of the essential physical components and published structural data, (II) computational modules that use the knowledge base to build or analyze structural models, and (III) a web-based user interface that supports multiple users, sessions and computations. We have built a prototype of Riboweb, and have used it to refine a rough model of the central domain of the 30S subunit from E. coli. procedure. Our results suggest that sophisticated and integrated computational capabilities can be delivered to biologists using this simple three-component architecture.

RNA secondary structure as a reusable interface to biological information resources

The dissemination of biological information has become critically dependent on the Internet and World Wide Web (WWW), which enable distributed access to information in a platform independent manner. The mode of interaction between biologists and on-line information resources, however, has been mostly limited to simple interface technologies such has hypertext links, tables and forms. The introduction of platform-independent runtime environments facilitates the development of more sophisticated WWW-based user interfaces. Until recently, most such interfaces have been tightly coupled to the underlying computation engines, and not separated as reusable components. We believe that many subdisciplines of biology have intuitive and familiar graphical representations of knowledge that can serve as multipurpose user interface elements. We call such graphical idioms "domain graphics". In order to illustrate the power of such graphics, we have built a reusable interface based on the standard two dimensional (2D) layout of RNA secondary structure. The interface can be used to represent any pre-computed layout of RNA, and takes as a parameters the sets of actions to be performed as a user interacts with the interface. It can provide to any associated application program information about the base, helix, or subsequence selected by the user. We show the versatility of this interface by using it as a special purpose interface to BLAST, Medline and the RNA MFOLD search/compute engines. These demonstrations are available at: http://www-smi.stanford.edu/projects/helix/pubs/ gene-combis-96/

Computational methods for defining the allowed conformational space of 16S rRNA based on chemical footprinting data

Structural models for 16S ribosomal RNA have been proposed based on combinations of crosslinking, chemical protection, shape, and phylogenetic evidence. These models have been based for the most part on independent data sets and different sets of modeling assumptions. In order to evaluate such models meaningfully, methods are required to explicitly model the spatial certainty with which individual structural components are positioned by specific data sets. In this report, we use a constraint satisfaction algorithm to explicitly assess the location of the secondary structural elements of the 16S RNA, as well as the certainty with which these elements can be positioned. The algorithm initially assumes that these helical elements can occupy any position and orientation and then systematically eliminates those positions and orientations that do not satisfy formally parameterized interpretations of structural constraints. Using a conservative interpretation of the hydroxyl radical footprinting data, the positions of the ribosomal proteins as defined by neutron diffraction studies, and the secondary structure of 16S rRNA, the location of the RNA secondary structural elements can be defined with an average precision of 25 A (ranging from 12.8 to 56.3 A). The uncertainty in individual helix positions is both heterogeneous and dependent upon the number of constraints imposed on the helix. The topology of the resulting model is consistent with previous models based on independent approaches. The result of our computation is a conservative upper bound on the possible positions of the RNA secondary structural elements allowed by this data set, and provides a suitable starting point for refinement with other sources of data or different sets of modeling assumptions.

Constraining volume by matching the moments of a distance distribution

The problem of computing a molecular structure from a set of distances arises in the interpretation of NMR data as well as other experimental methods that yield distance information. Techniques for computing structures must find conformations consistent with the distance data. There are often other constraints on the structure that must be satisfied as well. One of the most problematic constraints is the constraint on the total volume occupied by the atoms. In this paper, we use the first two moments (mean and variance) of an estimated distance distribution to constrain the volume of a computed structure. We show that a probabilistic algorithm for matching the first two moments of the estimated distance distribution significantly improves the quality of the solution, especially when the distance information alone is not sufficient to define the structure precisely. We also show that our method is not sensitive to small errors in the estimates of mean and variance of the distance distribution. Finally, we demonstrate the use of this constraint in computing a low-resolution structure of the 30S prokaryotic ribosomal subunit. Quantitative analysis of our results allows us to assess the information content contained in constraints on volume, and to show that in some cases addition of a volume constraint adds information roughly equivalent to doubling the number of input distances. Our results also demonstrate the flexibility of probabilistic representations of structural constraints, and the importance of including volume information to constrain structural computations-especially in the case of sparse data.

Three-dimensional transformation of capsids associated with genome packaging in a bacterial virus

Three-dimensional structures of the empty procapsid and the mature capsid of the Salmonella bacteriophage P22 have been determined to a resolution of 28 A using electron cryomicroscopy and computer image processing. The coat subunits in both the structures are arranged as pentamers and hexamers on a T = 7 icosahedral lattice. The two structures display significant differences in shape, size and intersubunit interactions. The empty procapsid is spherical in contrast to the distinctly larger and polyhedral mature capsid. The empty procapsid structure exhibits holes at all the quasi sixfold positions that are absent in the mature capsid. These holes may be the exit ports for scaffolding subunits. Detailed comparisons of the two structures indicate that extensive structural changes take place during maturation in all seven quasi-equivalent subunits. These changes cause flattening of the icosahedral facets, capsid expansion and closing of the holes. This process results in a stable and impenetrable capsid that protects the bacterial genome.