Machine learning to detect the SINEs of cancer.

We previously described an approach called RealSeqS to evaluate aneuploidy in plasma cell-free DNA through the amplification of ~350,000 repeated elements with a single primer. We hypothesized that an unbiased evaluation of the large amount of sequencing data obtained with RealSeqS might reveal other differences between plasma samples from patients with and without cancer. This hypothesis was tested through the development of a machine learning approach called Alu Profile Learning Using Sequencing (A-PLUS) and its application to 7615 samples from 5178 individuals, 2073 with solid cancer and the remainder without cancer. Samples from patients with cancer and controls were prespecified into four cohorts used for model training, analyte integration, and threshold determination, validation, and reproducibility. A-PLUS alone provided a sensitivity of 40.5% across 11 different cancer types in the validation cohort, at a specificity of 98.5%. Combining A-PLUS with aneuploidy and eight common protein biomarkers detected 51% of the cancers at 98.9% specificity. We found that part of the power of A-PLUS could be ascribed to a single feature-the global reduction of AluS subfamily elements in the circulating DNA of patients with solid cancer. We confirmed this reduction through the analysis of another independent dataset obtained with a different approach (whole-genome sequencing). The evaluation of Alu elements may therefore have the potential to enhance the performance of several methods designed for the earlier detection of cancer.

The Origin of Highly Elevated Cell-Free DNA in Healthy Individuals and Patients with Pancreatic, Colorectal, Lung, or Ovarian Cancer.

Cell-free DNA (cfDNA) concentrations from patients with cancer are often elevated compared with those of healthy controls, but the sources of this extra cfDNA have never been determined. To address this issue, we assessed cfDNA methylation patterns in 178 patients with cancers of the colon, pancreas, lung, or ovary and 64 patients without cancer. Eighty-three of these individuals had cfDNA concentrations much greater than those generally observed in healthy subjects. The major contributor of cfDNA in all samples was leukocytes, accounting for ∼76% of cfDNA, with neutrophils predominating. This was true regardless of whether the samples were derived from patients with cancer or the total plasma cfDNA concentration. High levels of cfDNA observed in patients with cancer did not come from either neoplastic cells or surrounding normal epithelial cells from the tumor's tissue of origin. These data suggest that cancers may have a systemic effect on cell turnover or DNA clearance. SIGNIFICANCE: The origin of excess cfDNA in patients with cancer is unknown. Using cfDNA methylation patterns, we determined that neither the tumor nor the surrounding normal tissue contributes this excess cfDNA-rather it comes from leukocytes. This finding suggests that cancers have a systemic impact on cell turnover or DNA clearance. See related commentary by Thierry and Pisareva, p. 2122. This article is featured in Selected Articles from This Issue, p. 2109.

Seq-ing the SINEs of central nervous system tumors in cerebrospinal fluid.

It is often challenging to distinguish cancerous from non-cancerous lesions in the brain using conventional diagnostic approaches. We introduce an analytic technique called Real-CSF (repetitive element aneuploidy sequencing in CSF) to detect cancers of the central nervous system from evaluation of DNA in the cerebrospinal fluid (CSF). Short interspersed nuclear elements (SINEs) are PCR amplified with a single primer pair, and the PCR products are evaluated by next-generation sequencing. Real-CSF assesses genome-wide copy-number alterations as well as focal amplifications of selected oncogenes. Real-CSF was applied to 280 CSF samples and correctly identified 67% of 184 cancerous and 96% of 96 non-cancerous brain lesions. CSF analysis was considerably more sensitive than standard-of-care cytology and plasma cell-free DNA analysis in the same patients. Real-CSF therefore has the capacity to be used in combination with other clinical, radiologic, and laboratory-based data to inform the diagnosis and management of patients with suspected cancers of the brain.

Detection of rare mutations, copy number alterations, and methylation in the same template DNA molecules.

The analysis of cell-free DNA (cfDNA) from plasma offers great promise for the earlier detection of cancer. At present, changes in DNA sequence, methylation, or copy number are the most sensitive ways to detect the presence of cancer. To further increase the sensitivity of such assays with limited amounts of sample, it would be useful to be able to evaluate the same template molecules for all these changes. Here, we report an approach, called MethylSaferSeqS, that achieves this goal, and can be applied to any standard library preparation method suitable for massively parallel sequencing. The innovative step was to copy both strands of each DNA-barcoded molecule with a primer that allows the subsequent separation of the original strands (retaining their 5-methylcytosine residues) from the copied strands (in which the 5-methylcytosine residues are replaced with unmodified cytosine residues). The epigenetic and genetic alterations present in the DNA molecules can then be obtained from the original and copied strands, respectively. We applied this approach to plasma from 265 individuals, including 198 with cancers of the pancreas, ovary, lung, and colon, and found the expected patterns of mutations, copy number alterations, and methylation. Furthermore, we could determine which original template DNA molecules were methylated and/or mutated. MethylSaferSeqS should be useful for addressing a variety of questions relating genetics and epigenetics.

The rapid and highly parallel identification of antibodies with defined biological activities by SLISY.

The therapeutic applications of antibodies are manifold and the emergence of SARS-CoV-2 provides a cogent example of the value of rapidly identifying biologically active antibodies. We describe an approach called SLISY (Sequencing-Linked ImmunoSorbent assaY) that in a single experiment can assess the binding specificity of millions of clones, be applied to any screen that links DNA sequence to a potential binding moiety, and requires only a single round of biopanning. We demonstrate this approach using an scFv library applied to cellular and protein targets to identify specific or broadly reacting antibodies. For a cellular target, we use paired HLA knockout cell lines to identify a panel of antibodies specific to HLA-A3. For a protein target, SLISY identifies 1279 clones that bound to the Receptor Binding Domain of the SARS-CoV-2 spike protein, with >40% of tested clones also neutralizing its interaction with ACE2 in in vitro assays. Using a multi-comparison SLISY against the Beta, Gamma, and Delta variants, we recovered clones that exhibited broad-spectrum neutralizing potential in vitro. By evaluating millions of scFvs simultaneously against multiple targets, SLISY allows the rapid identification of candidate scFvs with defined binding profiles facilitating the identification of antibodies with the desired biological activity.

Chemoradiation therapy alters the PD-L1 score in locoregional recurrent squamous cell carcinomas of the head and neck.

PD-L1 testing guides therapeutic decision-making for head and neck squamous cell carcinoma (HNSCC). We sought to understand whether chemoradiation therapy (CRT) influences the PD-L1 combined positive score (CPS) and other biomarkers of response to immunotherapy. PD-L1 expression was assessed using immunohistochemistry, and bulk RNA sequencing was performed on 146 HNSCC patients (65 primary sites, 50 paired local recurrences, and 31 paired regional recurrences). PD-L1 was scored using the CPS of ≥1, ≥20, and ≥50. Overall, 98 %, 54 %, and 17 % of HNSCCs had a CPS ≥1, ≥20, and ≥50, respectively. When using a cut-off of ≥1, CRT did not significantly change CPS at the locoregional recurrent site. However, there were significant changes when using CPS ≥20 or ≥50. The CPS changed for 32 % of patients when using a CPS ≥20 (p < 0.001). When using a CPS ≥50, there was a 20-23 % (p = 0.0058-0.00067) discordance rate at the site of locoregional recurrence. Oral cavity cancers had a significantly higher discordant rate than other primary sites for CPS ≥50, 44 % (8/18, p = 0.0058) and 58 % (7/12, p = 0.00067) discordance at the site of local and regional recurrence, respectively. When evaluating the 18 gene IFN-É£ signature predictive of response to anti-PD-1 blockade, there was a statistically significant increase in the IFN-É£ signature in recurrent larynx cancer (p = 0.02). Our study demonstrates that when using a higher cut-off of CPS ≥20 and ≥50, a repeat biopsy may be warranted after CRT for local and regional recurrent HNSCCs.

Detection of malignant peripheral nerve sheath tumors in patients with neurofibromatosis using aneuploidy and mutation identification in plasma.

Malignant peripheral nerve sheath tumors (MPNST) are the deadliest cancer that arises in individuals diagnosed with neurofibromatosis and account for nearly 5% of the 15,000 soft tissue sarcomas diagnosed in the United States each year. Comprised of neoplastic Schwann cells, primary risk factors for developing MPNST include existing plexiform neurofibromas (PN), prior radiotherapy treatment, and expansive germline mutations involving the entire NF1 gene and surrounding genes. PN develop in nearly 30-50% of patients with neurofibromatosis type 1 (NF1) and most often grow rapidly in the first decade of life. One of the most important aspects of clinical care for NF1 patients is monitoring PN for signs of malignant transformation to MPNST that occurs in 10-15% of patients. We perform aneuploidy analysis on ctDNA from 883 ostensibly healthy individuals and 28 patients with neurofibromas, including 7 patients with benign neurofibroma, 9 patients with PN and 12 patients with MPNST. Overall sensitivity for detecting MPNST using genome wide aneuploidy scoring was 33%, and analysis of sub-chromosomal copy number alterations (CNAs) improved sensitivity to 50% while retaining a high specificity of 97%. In addition, we performed mutation analysis on plasma cfDNA for a subset of patients and identified mutations in NF1, NF2, RB1, TP53BP2, and GOLGA2. Given the high throughput and relatively low sequencing coverage required by our assay, liquid biopsy represents a promising technology to identify incipient MPNST.

Comparison of next generation sequencing, droplet digital PCR, and quantitative real-time PCR for the earlier detection and quantification of HPV in HPV-positive oropharyngeal cancer.

BACKGROUND: Human papillomavirus (HPV) causes nearly 80% of oropharynx cancers diagnosed in the United States, with incidence increasing each year. Analysis of cfDNA in plasma and oral rinse has the potential to detect these cases earlier than their typical presentation, but their utility and the best method to detect HPV in plasma and oral rinse samples is unknown. MATERIALS AND METHODS: We directly compared next generation sequencing (NGS), droplet digital PCR (ddPCR), and quantitative real-time PCR (qPCR) for their ability to detect HPV16 DNA in plasma and oral rinse from 66 patients diagnosed with HPV16-positive oropharyngeal cancer (HPV16-OPC). RESULTS: HPV DNA detection by NGS and ddPCR in plasma samples both had good sensitivity (70%) for HPV16-OPC compared to 20.6% sensitivity by qPCR (p < 0.001). In oral rinse, NGS demonstrated a superior sensitivity of 75.0% as compared to both ddPCR (8.3%, p < 0.001) and qPCR (2.1%, p < 0.001). In a limited cohort of follow up patients, HPV levels detected in plasma by NGS but not ddPCR or qPCR reflected disease remission or progression. CONCLUSIONS: These results suggest that NGS has the best sensitivity for detecting HPV in both plasma and oral rinse and may play a role in monitoring patients for disease recurrence. Additional studies are needed to define the specificity of NGS for similar patient cohorts.

Myeloid Cells Are Enriched in Tonsillar Crypts, Providing Insight into the Viral Tropism of Human Papillomavirus.

Viruses are the second leading cause of cancer worldwide, and human papillomavirus (HPV)-associated head and neck cancers are increasing in incidence in the United States. HPV preferentially infects the crypts of the tonsils rather than the surface epithelium. The present study sought to characterize the unique microenvironment within the crypts to better understand the viral tropism of HPV to a lymphoid-rich organ. Laser-capture microdissection of distinct anatomic areas (crypts, surface epithelium, and germinal centers) of the tonsil, coupled with transcriptional analysis and multiparameter immunofluorescence staining demonstrated that the tonsillar crypts are enriched with myeloid populations that co-express multiple canonical and noncanonical immune checkpoints, including PD-L1, CTLA-4, HAVCR2 (TIM-3), ADORA2A, IDO1, BTLA, LGALS3, CDH1, CEACAM1, PVR, and C10orf54 (VISTA). The resident monocytes may foster a permissive microenvironment that facilitates HPV infection and persistence. Furthermore, the myeloid populations within HPV-associated tonsil cancers co-express the same immune checkpoints, providing insight into potential novel immunotherapeutic targets for HPV-associated head and neck cancers.

The mutational landscape of spinal chordomas and their sensitive detection using circulating tumor DNA.

BACKGROUND: Chordomas are the most common primary spinal column malignancy in the United States. The aim of this study was to determine whether chordomas may be detected by evaluating mutations in circulating tumor DNA (ctDNA). METHODS: Thirty-two patients with a biopsy-confirmed diagnosis of chordoma had blood drawn pre-operatively and/or at follow-up appointments. Mutations in the primary tumor were identified by whole exome sequencing and liquid biopsy by ddPCR and/or RACE-Seq was used to detect one or more of these mutations in plasma ctDNA at concurrent or later time points. RESULTS: At the time of initial blood draw, 87.1% of patients were ctDNA positive (P <.001). Follow-up blood draws in twenty of the patients suggest that ctDNA levels may reflect the clinical status of the disease. Patients with positive ctDNA levels were more likely to have greater mutant allele frequencies in their primary tumors (P = .004) and undergo radiotherapy (P = .02), and the presence of ctDNA may correlate with response to systemic chemotherapy and/or disease recurrence. CONCLUSIONS: Detection of ctDNA mutations may allow for the detection and monitoring of disease progression for chordomas.

Mebendazole and temozolomide in patients with newly diagnosed high-grade gliomas: results of a phase 1 clinical trial.

BACKGROUND: Mebendazole is an anthelmintic drug introduced for human use in 1971 that extends survival in preclinical models of glioblastoma and other brain cancers. METHODS: A single-center dose-escalation and safety study of mebendazole in 24 patients with newly diagnosed high-grade gliomas in combination with temozolomide was conducted. Patients received mebendazole in combination with adjuvant temozolomide after completing concurrent radiation plus temozolomide. Dose-escalation levels were 25, 50, 100, and 200 mg/kg/day of oral mebendazole. A total of 15 patients were enrolled at the highest dose studied of 200 mg/kg/day. Trough plasma levels of mebendazole were measured at 4, 8, and 16 weeks. RESULTS: Twenty-four patients (18 glioblastoma and 6 anaplastic glioma) were enrolled with a median age of 49.8 years. Four patients (at 200 mg/kg) developed elevated grade 3 alanine aminotransferase (ALT) and/or aspartate transaminase (AST) after 1 month, which reversed with lower dosing or discontinuation. Plasma levels of mebendazole were variable but generally increased with dose. Kaplan-Meier analysis showed a 21-month median overall survival with 41.7% of patients alive at 2 years and 25% at 3 and 4 years. Median progression-free survival (PFS) from the date of diagnosis for 17 patients taking more than 1 month of mebendazole was 13.1 months (95% confidence interval [CI]: 8.8-14.6 months) but for 7 patients who received less than 1 month of mebendazole PFS was 9.2 months (95% CI: 5.8-13.0 months). CONCLUSION: Mebendazole at doses up to 200 mg/kg demonstrated long-term safety and acceptable toxicity. Further studies are needed to determine mebendazole's efficacy in patients with malignant glioma.

The potential of cerebrospinal fluid-based liquid biopsy approaches in CNS tumors.

Cerebrospinal fluid (CSF) may be the best hope for minimally invasive diagnosis and treatment monitoring of central nervous system (CNS) malignancies. Discovery/validation of cell-free nucleic acid and protein biomarkers has the potential to revolutionize CNS cancer care, paving the way for presurgical evaluation, earlier detection of recurrence, and the selection of targeted therapies. While detection of mutations, changes in RNA and miRNA expression, epigenetic alterations, and elevations of protein levels have been detected in the CSF of patients with CNS tumors, most of these biomarkers remain unvalidated. In this review, we focus on the molecular changes that have been identified in a variety of CNS tumors and profile the approaches used to detect these alterations in clinical samples. We further emphasize the importance of systemic collection of CSF and the establishment of standardized collection protocols that will lead to better cross-study biomarker validation and hopefully FDA-approved clinical markers.

Applications of liquid biopsies for cancer.

Liquid biopsies have the potential to detect, characterize, and monitor cancers earlier than is possible with conventional approaches.

Detection and localization of surgically resectable cancers with a multi-analyte blood test.

Earlier detection is key to reducing cancer deaths. Here, we describe a blood test that can detect eight common cancer types through assessment of the levels of circulating proteins and mutations in cell-free DNA. We applied this test, called CancerSEEK, to 1005 patients with nonmetastatic, clinically detected cancers of the ovary, liver, stomach, pancreas, esophagus, colorectum, lung, or breast. CancerSEEK tests were positive in a median of 70% of the eight cancer types. The sensitivities ranged from 69 to 98% for the detection of five cancer types (ovary, liver, stomach, pancreas, and esophagus) for which there are no screening tests available for average-risk individuals. The specificity of CancerSEEK was greater than 99%: only 7 of 812 healthy controls scored positive. In addition, CancerSEEK localized the cancer to a small number of anatomic sites in a median of 83% of the patients.

PD-1 Expression in Head and Neck Squamous Cell Carcinomas Derives Primarily from Functionally Anergic CD4+ TILs in the Presence of PD-L1+ TAMs.

Oral tongue squamous cell carcinoma (OTSCC) is the most common oral cavity tumor. In this study, we examined the basis for the activity of programmed cell death protein (PD-1)-based immune checkpoint therapy that is being explored widely in head and neck cancers. Using multispectral imaging, we systematically investigated the OTSCC tumor microenvironment (TME) by evaluating the frequency of PD-1 expression in CD8+, CD4+, and FoxP3+ tumor-infiltrating lymphocytes (TIL). We also defined the cellular sources of PD-1 ligand (PD-L1) to evaluate the utility of PD-1:PD-L1 blocking antibody therapy in this patient population. PD-L1 was expressed in 79% of the OTSCC specimens examined within the TME. Expression of PD-L1 was associated with moderate to high levels of CD4+ and CD8+ TILs. We found that CD4+ TILs were present in equal or greater frequencies than CD8+ TILs in 94% of OTSCC and that CD4+FOXP3neg TILs were colocalized with PD-1/PD-L1/CD68 more frequently than CD8+ TILs. Both CD4+PD1+ and CD8+PD1+ TILs were anergic in the setting of PD-L1 expression. Overall, our results highlight the importance of CD4+ TILs as pivotal regulators of PD-L1 levels and in determining the responsiveness of OTSCC to PD1-based immune checkpoint therapy. Cancer Res; 77(22); 6365-74. ©2017 AACR.

Cancer-Associated Mutations in Endometriosis without Cancer.

BACKGROUND: Endometriosis, defined as the presence of ectopic endometrial stroma and epithelium, affects approximately 10% of reproductive-age women and can cause pelvic pain and infertility. Endometriotic lesions are considered to be benign inflammatory lesions but have cancerlike features such as local invasion and resistance to apoptosis. METHODS: We analyzed deeply infiltrating endometriotic lesions from 27 patients by means of exomewide sequencing (24 patients) or cancer-driver targeted sequencing (3 patients). Mutations were validated with the use of digital genomic methods in microdissected epithelium and stroma. Epithelial and stromal components of lesions from an additional 12 patients were analyzed by means of a droplet digital polymerase-chain-reaction (PCR) assay for recurrent activating KRAS mutations. RESULTS: Exome sequencing revealed somatic mutations in 19 of 24 patients (79%). Five patients harbored known cancer driver mutations in ARID1A, PIK3CA, KRAS, or PPP2R1A, which were validated by Safe-Sequencing System or immunohistochemical analysis. The likelihood of driver genes being affected at this rate in the absence of selection was estimated at P=0.001 (binomial test). Targeted sequencing and a droplet digital PCR assay identified KRAS mutations in 2 of 3 patients and 3 of 12 patients, respectively, with mutations in the epithelium but not the stroma. One patient harbored two different KRAS mutations, c.35G→T and c.35G→C, and another carried identical KRAS c.35G→A mutations in three distinct lesions. CONCLUSIONS: We found that lesions in deep infiltrating endometriosis, which are associated with virtually no risk of malignant transformation, harbor somatic cancer driver mutations. Ten of 39 deep infiltrating lesions (26%) carried driver mutations; all the tested somatic mutations appeared to be confined to the epithelial compartment of endometriotic lesions.

Bisulfite-converted duplexes for the strand-specific detection and quantification of rare mutations.

The identification of mutations that are present at low frequencies in clinical samples is an essential component of precision medicine. The development of molecular barcoding for next-generation sequencing has greatly enhanced the sensitivity of detecting such mutations by massively parallel sequencing. However, further improvements in specificity would be useful for a variety of applications. We herein describe a technology (BiSeqS) that can increase the specificity of sequencing by at least two orders of magnitude over and above that achieved with molecular barcoding and can be applied to any massively parallel sequencing instrument. BiSeqS employs bisulfite treatment to distinguish the two strands of molecularly barcoded DNA; its specificity arises from the requirement for the same mutation to be identified in both strands. Because no library preparation is required, the technology permits very efficient use of the template DNA as well as sequence reads, which are nearly all confined to the amplicons of interest. Such efficiency is critical for clinical samples, such as plasma, in which only tiny amounts of DNA are often available. We show here that BiSeqS can be applied to evaluate transversions, as well as small insertions or deletions, and can reliably detect one mutation among >10,000 wild-type molecules.

Ki-67 is required for maintenance of cancer stem cells but not cell proliferation.

Ki-67 expression is correlated with cell proliferation and is a prognostic marker for various cancers; however, its function is unknown. Here we demonstrate that genetic disruption of Ki-67 in human epithelial breast and colon cancer cells depletes the cancer stem cell niche. Ki-67 null cells had a proliferative disadvantage compared to wildtype controls in colony formation assays and displayed increased sensitivity to various chemotherapies. Ki-67 null cancer cells showed decreased and delayed tumor formation in xenograft assays, which was associated with a reduction in cancer stem cell markers. Immunohistochemical analyses of human breast cancers revealed that Ki-67 expression is maintained at equivalent or greater levels in metastatic sites of disease compared to matched primary tumors, suggesting that maintenance of Ki-67 expression is associated with metastatic/clonogenic potential. These results elucidate Ki-67's role in maintaining the cancer stem cell niche, which has potential diagnostic and therapeutic implications for human malignancies.

Identification of the murine H-2D(b) and human HLA-A*0201 MHC class I-restricted HPV6 E7-specific cytotoxic T lymphocyte epitopes.

Recurrent respiratory papillomatosis is caused by human papillomavirus (HPV) infection, most commonly types 6 (HPV-6) and 11 (HPV-11). Due to failed host immune responses, HPV is unable to be cleared from the host, resulting in recurrent growth of HPV-related lesions that can obstruct the lumen of the airway within the upper aerodigestive tract. In our murine model, the HPV-6b and HPV-11 E7 antigens are not innately immunogenic. In order to enhance the host immune responses against the HPV E7 antigen, we linked calreticulin (CRT) to HPV-6b E7 and found that vaccinating C57BL/6 mice with the HPV-6b CRT/E7 DNA vaccine is able to induce a CD8+ T cell response that recognizes an H-2D(b)-restricted E7aa21-29 epitope. Additionally, vaccination of HLA-A*0201 transgenic mice with HPV-6b CRT/E7 DNA generated a CD8+ T cell response against the E7aa82-90 epitope that was not observed in the wild-type C57BL/6 mice, indicating this T cell response is restricted to HLA-A*0201. In vivo cytotoxic T cell killing assays demonstrated that the vaccine-induced CD8+ T cells are able to efficiently kill target cells. Interestingly, the H-2D(b)-restricted E7aa21-29 sequence and the HLA-A*0201-restricted E7aa82-90 sequence are conserved between HPV-6b and HPV-11 and may represent shared immunogenic epitopes. The identification of the HPV-6b/HPV-11 CD8+ T cell epitopes facilitates the evaluation of various immunomodulatory strategies in preclinical models. More importantly, the identified HLA-A*0201-restricted T cell epitope may serve as a peptide vaccination strategy, as well as facilitate the monitoring of vaccine-induced HPV-specific immunologic responses in future human clinical trials.