Search

Targeting of the PD1/PD-L1 immune checkpoint pathway has rapidly gained acceptance as a therapeutic strategy for a growing number of malignancies. Testing for expression of PD-L1 in tumor cells and immune cells has been used as a companion or complementary test for drugs targeting the PD1/PD-L1 pathway. We evaluated the results of PD-L1 testing in a large reference lab cohort. Using Food and Drug Administration-approved methods and interpretive instructions for each individual test, 62,896 cases were evaluated for PD-L1 using antibody clone 22C3, 28-8, SP142, or SP263. Case data analyzed included test results and information on tumor location and clinical history. No clinical outcome information was available and no attempt was made to correlate PD-L1 results with any other tests performed. The following numbers of cases were evaluated: 22C3 with tumor proportion score [n = 52585], 22C3 with combined positive score [n = 2631], 28-8 [n = 4191], SP142 [n = 850], and SP263 [n = 70]. In 22C3/tumor proportion score cases, the general results were as follows: negative 33.1% (n = 17,405), (low) expression 33.9% (n = 17,822), and high expression 29.5% (n = 15,486). In cases identified as metastatic, the results were as follows: negative 35.9% (n = 1411), (low) expression 30.8% (n = 1211), and high expression 30.7% (n = 1208). We found broad ranges of expression in tumor types with increasing positivity, as adenocarcinomas were reported as poorly differentiated, whereas squamous cell carcinomas showed more positivity as tumors were described as well-differentiated. The results of many individual tumor types were evaluated and showed, in general, high levels of positive expression. Practical challenges and observations of PD-L1 stain results and interpretation are also discussed.

AimsAt the time of analysis, two widely used, drug-specific, tumour-cell programmed death ligand 1 (PD-L1) assays were approved by the US Food and Drug Administration for anti-PD-1 therapies: the Dako PD-L1 immunohistochemistry (IHC) 28-8 pharmDx assay and the Dako PD-L1 IHC 22C3 pharmDx assay. Given that the majority of current PD-L1 testing in US clinical practice is performed at commercial reference laboratories, we aimed to evaluate the concordance of the 28-8 and 22C3 assays in a real-world setting.MethodsMatched PD-L1 IHC 28-8 and 22C3 results from routine assessment were obtained from 1930 patients, including 412 confirmed to have lung cancer, submitted from hospitals in over 38 US states/territories. Biopsies were stained, reviewed and scored by trained/certified pathologists at a single cancer reference laboratory between 2015 and 2017. Rate of concordance between assay findings was assessed by Bland-Altman analysis; overall per cent agreement (OPA), positive per cent agreement and negative per cent agreement; and Cohen’s kappa.ResultsPD-L1 IHC 28-8 and 22C3 displayed strong correlation across all samples and in samples with a confirmed lung cancer diagnosis irrespective of biopsy site. The OPA was 97%–98% for all samples, depending on the expression level defining PD-L1 positivity. In the Bland-Altman analysis, the mean difference in percentage of tumour cells positively stained for PD-L1 between the paired assay findings was –0.80% for all samples and –0.93% in samples with a confirmed lung cancer diagnosis.ConclusionsThese data, in conjunction with recent findings, support the analytical concordance of the PD-L1 IHC 28-8 and 22C3 assays for assessing per cent tumour-cell membrane PD-L1 expression.

Purpose While HER2 testing is well established in directing appropriate treatment for breast cancer, a small percentage of cases show equivocal results by immunohistochemistry (IHC) and fluorescence in situ hybridization (FISH). Alternative probes may be used in equivocal cases. We present a single community-based institution’s experience in further evaluating these cases. Patients and methods Between 2014 and 2016, 4255 samples were submitted for HER2 amplification testing by alternative probes, TP53, RAI1, and RARA. Of the patients tested by FISH, 505/3908 (12.9%) also had IHC data. Results Most (73.9%) FISH equivocal cases remained equivocal after IHC testing. However, 50.5% of equivocal cases were classified as HER2 amplified by alternative probes. Most cases were positive by more than one probe: 78% of positive cases by RAI1 and 73.9% by TP53. There was a significant difference between IHC and FISH alternative testing (p

// Maher Albitar 1 , Sucha Sudarsanam 1 , Wanlong Ma 1 , Shiping Jiang 1 , Wayne Chen 1 , Vincent Funari 1 , Forrest Blocker 1 and Sally Agersborg 1 1 NeoGenomics Laboratories, Aliso Viejo, CA, USA Correspondence to: Maher Albitar, email: maher.albitar@neogenomics.com Keywords: NSCLC; PD-L1; EGFR; MET; TP53 Received: October 06, 2017 Accepted: January 09, 2018 Published: February 08, 2018 ABSTRACT Background: The role of MET amplification in lung cancer, particularly in relation to checkpoint inhibition and EGFR WT, has not been fully explored. In this study, we correlated PD-L1 expression with MET amplification and EGFR , KRAS , or TP53 mutation in primary lung cancer. Methods: In this retrospective study, tissue collected from 471 various tumors, including 397 lung cancers, was tested for MET amplification by FISH with a MET /centromere probe. PD-L1 expression was evaluated using clone SP142 and standard immunohistochemistry, and TP53 , KRAS , and EGFR mutations were tested using next generation sequencing. Results: Our results revealed that PD-L1 expression in non-small cell lung cancer is inversely correlated with EGFR mutation (P=0.0003), and positively correlated with TP53 mutation (P=0.0001) and MET amplification (P=0.004). Patients with TP53 mutations had significantly higher MET amplification (P=0.007), and were more likely (P=0.0002) to be EGFR wild type. There was no correlation between KRAS mutation and overall PD-L1 expression, but significant positive correlation between PD-L1 expression and KRAS with TP53 co-mutation (P=0.0002). A cut-off for the ratio of MET : centromere signal was determined as 1.5%, and 4% of lung cancer patients were identified as MET amplified. Conclusions: This data suggests that in lung cancer both MET and TP53 play direct roles in regulating PD-L1 opposing EGFR . Moreover, KRAS and TP53 co-mutation may cooperate to drive PD-L1 expression in lung cancer. Adding MET or TP53 inhibitors to checkpoint inhibitors may be an attractive combination therapy in patients with lung cancer and MET amplification.

Background The role of MET amplification in lung cancer, particularly in relation to checkpoint inhibition and EGFR WT, has not been fully explored. In this study, we correlated PD-L1 expression with MET amplification and EGFR, KRAS, or TP53 mutation in primary lung cancer. Methods In this retrospective study, tissue collected from 471 various tumors, including 397 lung cancers, was tested for MET amplification by FISH with a MET/centromere probe. PD-L1 expression was evaluated using clone SP142 and standard immunohistochemistry, and TP53, KRAS, and EGFR mutations were tested using next generation sequencing. Results Our results revealed that PD-L1 expression in non-small cell lung cancer is inversely correlated with EGFR mutation (P=0.0003), and positively correlated with TP53 mutation (P=0.0001) and MET amplification (P=0.004). Patients with TP53 mutations had significantly higher MET amplification (P=0.007), and were more likely (P=0.0002) to be EGFR wild type. There was no correlation between KRAS mutation and overall PD-L1 expression, but significant positive correlation between PD-L1 expression and KRAS with TP53 co-mutation (P=0.0002). A cut-off for the ratio of MET: centromere signal was determined as 1.5%, and 4% of lung cancer patients were identified as MET amplified. Conclusions This data suggests that in lung cancer both MET and TP53 play direct roles in regulating PD-L1 opposing EGFR. Moreover, KRAS and TP53 co-mutation may cooperate to drive PD-L1 expression in lung cancer. Adding MET or TP53 inhibitors to checkpoint inhibitors may be an attractive combination therapy in patients with lung cancer and MET amplification.

Background Clinical trials show favorable OS for patients (pts) with incident aNSCLC and programmed death ligand 1 (PD-L1) expression treated with PD-L1 immune checkpoint inhibitor (ICI) pembrolizumab (PEM) or PEM-chemotherapy compared to no ICI. We use real world data to examine OS in pts with PEM-based regimens vs NoICI as initial aNSCLC treatment by PD-L1 expression. Methods This retrospective study used de-identified administrative claims, including month of death (MoD) from regular operations, spanning 1/2017-12/2018 from a large, national US insurer. Claims were linked with biomarker test data from a CLIA-certified cancer diagnostics lab. During 2018, Medicare Advantage pts having ≥1 claim with diagnosis (dx) of lung cancer (ICD-10 C34), advanced disease (C77-C79), a PD-L1 test (CPT 88360) and test result, Tumor Proportion Score (TPS), were included. Pts were newly diagnosed (no C34 in 2017) and continuously enrolled in 2017 & 2018 or until death in 2018. Cohorts were based on initial therapy as PEM-based or NoICI and TPS result. Pts were followed from dx to MoD or 12/2018. OS was MoD, minus month of dx; otherwise censored at 12/2018. Multivariable Cox proportional hazard regressions compared OS across cohorts. Results 457 pts met study criteria: 126 PEM-based and 331 NoICI. Median age was 76 years in both cohorts. PEM-based had more males (63% vs 51%; p = 0.02) Median months followed was 6 (PEM-based) and 4 (NoICI). For PEM-based , 23 (18%) had TPS 0, 36 (29%) had TPS 1- 49 and 67 (53%) had TPS 50+. For NoICI, 118 (36%) had TPS 0, 120 (36%) had TPS 1- 49, and 93 (28%) had TPS 50+. 20% (n = 25) of PEM-based and 31% (n = 104) of NoICI died during the study period. Mortality hazard ratios for PEM-based vs NoICI by TPS score were: TPS 0, 0.46 (CI 95% 0.16 – 1.30); TPS 1 to 49, 0.73 (CI 95% 0.35 – 1.52); and TPS 50+,0.35 (CI 95% 0.18 – 0.68). OS was more favorable for the PEM-based cohort for all TPS groups, though only TPS 50+ was significant. Conclusion This study used administrative claims and lab test data allowing examination of effectiveness of PEM-based vs. NoICI therapy for aNSCLC. As in clinical trials, this analysis suggests OS benefit for PEM-based vs NoICI. Further analysis with longer follow-up will better assess OS benefit by TPS score. Legal entity responsible for the study UnitedHealth Group. Funding Has not received any funding. Disclosure J. Staib: Shareholder / Stockholder / Stock options, Full / Part-time employment: UnitedHealth Group; Shareholder / Stockholder / Stock options: Bristol Myers Squibb; Shareholder / Stockholder / Stock options: Merck; Shareholder / Stockholder / Stock options: Natera; Shareholder / Stockholder / Stock options: NeoGenomics. S. Sudarsanam: Shareholder / Stockholder / Stock options, Full / Part-time employment: NeoGenomics Laboratories. S. Dacosta Byfield: Shareholder / Stockholder / Stock options, Full / Part-time employment: UnitedHealth Group. C. Kennedy: Shareholder / Stockholder / Stock options, Full / Part-time employment: UnitedHealth Group. L. Weiss: Shareholder / Stockholder / Stock options, Full / Part-time employment: NeoGenomics Laboratories.

Genome-wide RNA splicing (with gene expression) can be used to discover variations that drive specific diseases and / or change the susceptibility in individuals to drug responses including tissue specific toxicities. Evidence linking causative SNPs to individual splicing differences between individuals is emerging and this may lead to a better understanding of susceptibilities related to rare drug-induced toxicities. The development of more sensitive genomics tools is expected to further the study of variations in molecular phenotype from alternative splicing of pre-mRNA. This report highlights a genomics platform developed to measure splicing changes that occur in response to drug exposures, and therefore is applicable for the study of drug-induced toxicity. The platform is applicable for humans, all toxicology species, and specialized model systems. For efficiency, multiple samples can be combined into a single sequencing run and individual sequences can be separated via informatics. Biobanked specimens from clinical trials, toxicology studies, from commercial sources, and/or from public 'omics' data resources such as in NCBI are the only sample or non-sample data requirements.

Background: Acute myeloid leukemia (AML) and myelodysplastic syndromes (MDS) are heterogeneous groups of disorders with a spectrum of clinical presentations and outcomes. Prognosis depends on various factors including age, karyotype, performance status, previous treatments and mutation status. Genetic profiling with next generation sequencing (NGS) is increasingly being used at diagnosis to detect presence of somatic mutations for prognostic risk stratification, and identification of therapeutic targets. Here we seek to identify epidemiologic differences in genetic mutations based on population and demographic data in patients with a preliminary diagnosis of AML or MDS. Methods: NGS mutation data were collected for 62 genes related to AML/MDS on a total of 10,934 patient samples submitted for testing for suspected AML/MDS. Samples were run on either the 54-gene NeoType Myeloid Disorders Profile (Neogenomics) or the 37-gene OnkoSight Myeloid Malignancies Panel (Genpath). The frequency of gene mutations (i.e., the number of patient samples with mutations for each gene) was identified for 58 counties in the USA. Counties in which fewer than 50 patient samples were tested were excluded from the dataset to minimize sampling bias. The counties were then grouped into 3 categories ranging from most urban to most rural based on a modified version of the 2013 National Center for Health Statistics classification system G-1 - > 1,000,000 (N=34), G2 - 250,000 - 1,000,000 (N=16) and G3 < 250,000 (N=8). One-way ANOVA and subsequent T-tests were performed for all genes based on the 3 urban-rural groupings to determine if significant differences in frequency of mutations exist between the 3 groupings. Difference of proportions tests were performed to identify variations in the patterns of frequency between counties. Results: The top 10 most frequent mutations were TET2, ASXL1, DNMT3A, SRSF2, TP53, RUNX1, SF3B1, U2AF1, NRAS, and NPM1(highest to lowest). The three mutations with the widest range of variability across counties were DNMT3A, TET2, and ASXL1 (DTA mutations). The median age across all counties was 68 (range 44-77). The county with the youngest and highest median age was Montgomery, CA and Pinellas, FL, with ASXL1 and TET2 as the most frequent mutation, respectively. Sacramento, CA had an unusually high rate of ASXL1 mutations (> 24%). ASXL1 was also significantly higher than TET2 in Essex, NJ; Montgomery, MD; and Sacramento, CA. (p=0.0287, p < 0.0001, p < 0.0001, respectively). IDH1was higher than IDH2 for Martin, FL (p= 0.0481). The highest frequency of TP53 mutation (24%) was in Bexar, TX compared to Montgomery, MD which had the lowest (2%). When counties were grouped based on population density, the frequency of RUNX1 and SF3B1 were statistically different across the counties, highest in G2 compared to G1 and G3 (p= 0.0294 and 0.0010 respectively). Conclusion: Our retrospective observational study is the first of its kind to look at genetic mutations in AML/MDS patients across the United States using commercially available NGS platforms. In general, patients had analogous combinations and frequencies of mutations commonly seen in AML and MDS, and the wide variation in frequency of DTA mutations is consistent with information known about age-related clonal hematopoiesis. In counties that showed a higher rate of ASXL1 > TET2, there may be a potential environmental factor accounting for this difference as the reverse is more commonly seen. Likewise, IDH1 mutations are typically seen at a lower frequency compared to IDH2, and it is interesting to note the reverse in Martin, FL (G2) despite the mutation frequency of all other genes being similar compared to the median for all counties. Our data analysis also showed a significant difference in frequency of mutations for TP53, RUNX1 and SF3B1. These variations have important implications in regard to prognosis, and the approach to treatment. Our observations suggest further investigation is warranted to explore potential environmental exposures related to somatic mutational patterns in patients with AML and MDS. Disclosures McCloskey: Jazz Pharmaceuticals: Consultancy, Speakers Bureau; Celgene Pharmaceuticals: Honoraria, Speakers Bureau; Amgen Pharmaceuticals: Speakers Bureau; Pfizer: Consultancy; Takeda Pharmaceuticals: Consultancy, Speakers Bureau; COTA: Equity Ownership. Koprivnikar:Amgen: Speakers Bureau; Otsuka: Consultancy; Alexion: Consultancy, Speakers Bureau.

e15604Background: Microsatellite instability (MSI) and DNA mismatch repair defect (dMMR) are dependable predictive biomarkers of colorectal cancer (CRC) response to checkpoint inhibitors. We explor...

3022Background: DNA mismatch repair deficiency (dMMR) can be tested by immunohistochemistry (IHC) or microsatellite instability (MSI). While either IHC and MSI is adequate for establishing Lynch sy...

3088Background: Predictive biomarkers for selecting of patients who may benefit or experience serious adverse effects from checkpoint blockade therapy are urgently needed. These biomarkers may vary...

We have determined the full protein kinase (PK) complement (kinome) of mouse. This set of 540 genes includes many novel kinases and corrections or extensions to >150 published sequences. The mouse has orthologs for 510 of the 518 human PKs. Nonorthologous kinases arise only by retrotransposition and gene decay. Orthologous kinase pairs vary in sequence conservation along their length, creating a map of functionally important regions for every kinase pair. Many species-specific sequence inserts exist and are frequently alternatively spliced, allowing for the creation of evolutionary lineage-specific functions. Ninety-seven kinase pseudogenes were found, all distinct from the 107 human kinase pseudogenes. Chromosomal mapping links 163 kinases to mutant phenotypes and unlocks the use of mouse genetics to determine functions of orthologous human kinases.

Protein phosphorylation controls many cellular processes, especially those involved in intercellular communication and coordination of complex functions. To explore the evolution of protein phosphorylation, we compared the protein kinase complements ('kinomes') of budding yeast, worm and fly, with known human kinases. We classify kinases into putative orthologous groups with conserved functions and discuss kinase families and pathways that are unique, expanded or lost in each lineage. Fly and human share several kinase families involved in immunity, neurobiology, cell cycle and morphogenesis that are absent from worm, suggesting that these functions might have evolved after the divergence of nematodes from the main metazoan lineage.

Background: Higher levels of PD-L1 expression at the surface of tumor cells have been associated with increased response to anti-PD-1 therapies (Santabarbara G et al. Ann Transl Med. 2016;4:215; Borghaei H et al. N Engl J Med. 2015;373:1627-1639; Brahmer J et al. N Engl J Med. 2015;373:123-135). The FDA has 2 approved tests associated with the use of these agents. The Dako PD-L1 IHC 22C3 kit is approved as a companion diagnostic for the use of pembrolizumab in the first- and second-line non-small-cell lung cancer (NSCLC) settings, where diagnostic testing is required for its use. The Dako PD-L1 IHC 28-8 kit is approved as a complementary diagnostic for nivolumab therapy in second-line NSCLC, where testing is not required for its use. NeoGenomics Laboratories stained a subset of the samples received for PD-L1 testing with both 22C3 and 28-8 on the same biopsy specimen. This dataset presents an opportunity to explore real-world concordance testing for 22C3 and 28-8 antibodies. Methods: The analysis was performed on a dataset obtained from Symphony Health Solutions describing the PD-L1 biomarker test results and annotations reported by NeoGenomics Laboratories between Oct 7, 2015, and Aug 31, 2016. A total of 5217 biomarker test results were available for 4528 patients afflicted by diverse malignancies. Concordance testing was performed on 556 unique patients that had a single 22C3/28-8 pairing collected from the same biopsy specimen. All staining was performed using the FDA-approved in vitro diagnostic assay using the 22C3 antibody or the 28-8 antibody on the Dako Link 48. Cases were reviewed and scored randomly across multiple pathologists, who received special training and certification in scoring. All statistical analyses were performed in SAS. Results: Paired biopsy specimens stained with both 22C3 and 28-8 displayed high degrees of correlation (Spearman’s correlation co-efficient = 0.97). In addition, Bland-Altman analysis revealed that the mean difference in the percentage of tumor cells positively stained for PD-L1 between the paired 22C3 and 28-8 assay findings was 0.48% [95% prediction limit (-12.41% to 13.37%)]. Across expression levels both antibodies displayed a similar probability of being interpreted as greater than one another (43 of 556 where 28-8>22C3 and 40 of 556 where 22C3>28-8), with no clear direction of a single antibody displaying greater sensitivity. When analyzed at the ≥ 1%, ≥ 5%, ≥ 10%, ≥ 25% and ≥ 50% thresholds, the overall rate of agreement was between 96.8% and 98.2% and associated with a level of agreement (Cohen’s kappa) between 0.92 and 0.96. Conclusion: Both the 22C3 and 28-8 diagnostic assays show strong agreement in a single central laboratory real-world setting. These data, in conjunction with recent findings from analytical comparability studies reported elsewhere, support the potential interchangeability of these assays for diagnostic interpretation. Citation Format: Cory Batenchuk, Maher Albitar, Sucha Sudarsanam, Vladislav Chizhevsky, Chelsea Jin, Virginia A. Burns. A comparative study of PD-L1 IHC 22C3 and 28-8 FDA-approved diagnostic assays in cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 4015. doi:10.1158/1538-7445.AM2017-4015

Introduction: MET gene activation has been reported to be associated with resistance to EGFR inhibitors in lung cancer. Resistance to EGFR inhibitors is also reported to be associated with upregulation of PD-L1. We studied PD-L1 expression levels with MET gene amplification and mutation in EGFR, KRAS, or TP53 in a large number of primary untreated lung cancers. Methods: Tissue samples collected from 397 core biopsies or resections from lung cancers were studied for MET gene amplification by fluorescent in-situ hybridization (FISH) using MET (7q31) probe and centromere 7 as a control. Signals were quantified. PD-L1 expression on the same samples was evaluated using clone SP142 and standard immunohistochemistry (IHC) procedure. Samples were also sequenced using next generation sequencing (NGS) for mutations in TP53, KRAS, and EGFR. Results: PD-L1 expression was detected in 166 patients (42%). Twenty seven (7%) had expression between 1% and 5%, 61 (15%) between 1% and 20%, 92 (23%) between 1% and 50%, and 105 (26%) had PD-L1 > 50% in tumor cells. Ratio of MET: centromere signals was > 1.5 in 16 (4%) of patients. Patients with MET ratio > 1.5% had significantly higher (P=0.004) percentage of PD-L1 as a continuous variable as well as when cut-off points of 5% (P=0.01), 20% (P=0.0006), and 50% (P=0.01) were used. Patients with EGFR mutation had significantly lower levels of PD-L1 expression (P=0.003). When a cut-off point of 50% is used for PD-L1 expression, the EGFR-mutant cases had significantly less number of positive cases (P=0.0003). There was no correlation between the presence of High MET: Ratio and EGFR mutation. There was no correlation between KRAS mutation and overall PD-L1 expression (P=0.4). There was no correlation between MET ratio and KRAS mutation. Patients with TP53 mutation had significantly higher MET ratio when a cut-off at 1.5 is used (P=0.01). Also there was significant correlation between TP53 mutation and overall PD-L1 expression (P=0.0001). This remained significant when cut-off point of 50% or 5% are used (P=0.0004 and P=0.007, respectively). TP53 mutation was significantly more common in EGFR wild-type cases (P=0.0002). Conclusions: Extra copies of MET gene as detected by FISH testing in therapy-naïve lung cancer patients is associated with higher expression of PD-L1, while EGFR-mutant lung cancers had significantly lower expression of PD-L1 when clone SP142 is used and NGS is used for detecting EGFR mutations. Patients with TP53 mutation had strikingly high expression of PD-L1 using SP142 clone and higher copy number of MET gene. This data suggests that in lung cancer, both MET and TP53 genes play a direct role in regulating PD-L1 expression opposing the EGFR gene, which appears to suppress PD-L1 expression. KRAS gene may not be involved in PD-L1 expression in lung cancer. Citation Format: Maher Albitar, Sucha Sudarsanam, Wanlong Ma, Shiping Jiang, Wayne Chen, Vincent Funari, Steven Brodie, Sally Agersborg. Correlation between MET gene amplification and TP53 mutation in upregulating PD-L1 expression in EGFR wild-type lung cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 5618. doi:10.1158/1538-7445.AM2017-5618

e14500 Background: PD-L1 expression as detected by immunohistochemistry (IHC) is significantly lower in colorectal cancers (CRC) when compared with lung cancer or other types of cancer. We explored if mutations in the RAS/RAF gene family, TP53 or PIK3CA can define a subgroup of CRC that express PD-L1. Methods: Tissue samples collected from 107 patients with CRC were studied for the expression of PD-L1 using clone SP142. The same samples were also tested for mutations in NRAS, KRAS, HRAS, BRAF, TP53, and PIK3CA using Next Generation Sequencing (NGS). Results: Of the 107 CRC samples only 15 (14%) showed PD-L1 positive tumor cells (≥1%) and 8 of the 15 (7.5% of total) had PD-L1 in ≤5% of tumor cells. Detected mutations in these samples were as follows: TP53 65%, KRAS 49.5%, PI3KCA 22.5%, NRAS 5%, HRAS 1%, and BRAF 17%. There was no correlation between PD-L1 expression and mutation status in any of the RAS/RAF genes. There was also no correlation between TP53 mutation and PD-L1 expression. This was true irrespective if PD-L1 expression is considered as a continuous variable or when cut-off points of 5%, 20%, or 50% were used. However, patients without any mutation in RAS or TP53 had significantly (P = 0.005) more expression of PD-L1 when cut-off point of 5% is used. This remained true if PD-L1 expression is considered as a continuous variable (P = 0.04). There was no correlation between PIK3CA and PD-L1 expression. Conclusions: PD-L1 expression is significantly more common in CRC that lack mutations in RAS or TP53. PD-L1 expression is detected in 31% of patients with wild-type RAS/TP53 as compared with 12% in patients with RAS/TP53 mutations (P = 0.04). If a cut-off point of 5% is used, 31% of RAS/TP53-wild-type CRC were positive for PD-L1, while only 6% of RAS/TP53- mutant CRC were positive for PD-L1 (P = 0.005). This suggests that in CRC without RAS/TP53 mutation, the PD-L1 may play a more important role in oncogenesis. Exploring immunotherapy in this group of CRC patients might be justified.

e21050 Background: BRAF mutation and PD-L1 expression appear independent in melanoma. We have established that PD-L1 clones 22C3 and 28.8 show almost identical results. In this study we correlated between BRAF mutation and PD-L1 expression as detected by 22C3/28.8 and SP142 in melanoma clinical samples. Methods: Melanoma samples were tested for PD-L1 expression and BRAF mutation. IHC testing for PD-L1 22C3 or 28.8 were tested using FDA-approved kits as recommended. Testing with SP142 (Spring Biosciences; LDT) was performed using standard techniques. BRAF testing in combination with SP142 PD-L1 testing was performed using next generation sequencing covering exons 11 and 15. Samples tested with 22C3 and 28.8 clones were tested using FDA-cleared BRAF kits (Cobas, Therascreen) testing V600 only. Results: 68 samples were tested for PD-L1 expression with clone 22C3, 67 with 28.8, and 56 with SP142. There was no overall statistical difference between the three clones in PD-L1 expression (P = 0.41). For combined 22C3 and 28.8, 61 of 135 cases (45%) had PD-L1 expression and 31 (23%) had BRAF mutation detected using FDA kits (V600). There was no statistical correlation between PD-L1 expression and BRAF mutation in this group (P = 0.9) at any cut-off. In cases tested with SP142 clone, BRAF mutation was detected in 31 cases (55%) by NGS, higher than using the FDA kit (P < 0.0001). In V600 mutations by NGS, 14 cases (25%) were positive, similar FDA kits. In SP142 cases, PD-L1 was positive in 45% of cases. There was no correlation between BRAF mutation and PD-L1 using SP142 expression as a continuous variable (P = 0.59) or cut-off points of 5% or 50%. When cut-off of 20% was used, significant inverse correlation with BRAF mutation (P = 0.004) was identified and was maintained if V600 codon only was evaluated. Conclusions: There is no correlation between BRAF mutation and PD-L1 expression as detected using clones 22C3 and 28.8. With SP142, a negative correlation between PD-L1 expression and BRAF mutation is identified with a 20% cut-off. The rate of BRAF mutations practically doubles when NGS is used and exons 11 and 15 are included in the testing, in contrast to FDA approved testing.

With the availability of the nearly complete genomic sequence of C. elegans, the first multicellular organism to be sequenced, molecular biology has definitely entered the postgenomic era. Annotation of the genomic sequence, which refers to identifying the genes and other biologically relevant sections of the genome, is an important and nontrivial next step. A first-pass annotation will be necessarily incomplete but will drive further biological experiments, which in turn will help to annotate the genome better. Given the scale of the genome sequence analysis, it is clear that the annotation should be automated as much as possible without sacrificing the quality of analysis. In this work, we outline our approach to identifying the protein kinases of C. elegans from the genomic sequence. We describe new tools we have developed for analysis, management and visualization of genomic data. By developing modular and scalable solutions, this study has provided a framework for future analysis of the Drosophila and human genomes.

Caenorhabditis elegans should soon be the first multicellular organism whose complete genomic sequence has been determined. This achievement provides a unique opportunity for a comprehensive assessment of the signal transduction molecules required for the existence of a multicellular animal. Although the worm C. elegans may not much resemble humans, the molecules that regulate signal transduction in these two organisms prove to be quite similar. We focus here on the content and diversity of protein kinases present in worms, together with an assessment of other classes of proteins that regulate protein phosphorylation. By systematic analysis of the 19,099 predicted C. elegans proteins, and thorough analysis of the finished and unfinished genomic sequences, we have identified 411 full length protein kinases and 21 partial kinase fragments. We also describe 82 additional proteins that are predicted to be structurally similar to conventional protein kinases even though they share minimal primary sequence identity. Finally, the richness of phosphorylation-dependent signaling pathways in worms is further supported with the identification of 185 protein phosphatases and 128 phosphoprotein-binding domains (SH2, PTB, STYX, SBF, 14-3-3, FHA, and WW) in the worm genome.

A method to sample the conformational space of polypeptides in a sequence-dependent manner was developed and applied to predict backbone conformations of polypeptides. The method uses a database of phi(i)+1 and psi(i) angles sorted into pools corresponding to 400 dimers of naturally occurring amino acids. The pools of phi(i)+1 and psi(i) angles for dimers were further dynamically divided into sub-pools based on the homology of two amino acids on either side of the dimer. Typically, 10000 backbone conformations for every overlapping hexamer of a polypeptide are generated by randomly assigning a set of phi(i)+1 and psi(i) angles for every dimer from the sorted sub-pool for that dimer. The conformational preference for each hexamer is evaluated by superimposing each simulated conformation on a set of pre-selected hexameric templates which are derived from high-resolution crystal structures representing helical, extended and turn conformations. A model is classified as compatible with a template if it has a backbone root mean square deviation of < or = 1 A to that template. The template that has the highest population of compatible models for a given hexamer is interpreted as the most probable conformation of that hexamer. The method was applied to predict backbone conformations that arise primarily from local interactions and the predictions were compared with experimental results. Results from simulations on structurally characterized polypeptides in helical, extended, turn and mixed alphabeta conformations are presented to demonstrate the predictive powers of the method. It is also demonstrated how models of overlapping hexamers can be assembled to obtain near-native models for the entire sequence.

We present an automated method for modeling backbones of protein loops. The method samples a database of phi i + 1 and psi i angles constructed from a nonredundant version of the Protein Data Bank (PDB). The dihedral angles phi i + 1 and psi i completely define the backbone conformation of a dimer when standard bond lengths, bond angles, and a trans planar peptide configuration are used. For the 400 possible dimers resulting from 20 natural amino acids, a list of allowed phi i + 1, psi i pairs for each dimer is created by pooling all such pairs from the loop segments of each protein in the nonredundant version of the PDB. Starting from the N-terminus of the loop sequence, conformations are generated by assigning randomly selected pairs of phi i + 1, psi i for each dimer from the respective pool using standard bond lengths, bond angles, and a trans peptide configuration. We use this database to simulate protein loops of lengths varying from 5 to 11 amino acids in five proteins of known three-dimensional structures. Typically, 10,000-50,000 models are simulated for each protein loop and are evaluated for stereochemical consistency. Depending on the length and sequence of a given loop, 50-80% of the models generated have no stereochemical strain in the backbone atoms. We demonstrate that, when simulated loops are extended to include flanking residues from homologous segments, only very few loops from an ensemble of sterically allowed conformations orient the flanking segments consistent with the protein topology. The presence of near-native backbone conformations for loops from five different proteins suggests the completeness of the dimeric database for use in modeling loops of homologous proteins. Here, we take advantage of this observation to design a method that filters near-native loop conformations from an ensemble of sterically allowed conformations. We demonstrate that our method eliminates the need for a loop-closure algorithm and hence allows for the use of topological constraints of the homologous proteins or disulfide constraints to filter near-native loop conformations.

Biomarkers are characteristics objectively measured and evaluated as indicators of: normal biologic processes, pathogenic processes, or pharmacologic response(s) to a therapeutic intervention. In environmental research and risk assessment, biomarkers are frequently referred to as indicators of human or environmental hazards. Discovering and implementing new biomarkers for toxicity caused by exposure to a chemical either from a therapeutic intervention or accidentally through the environment continues to be pursued through the use of animal models to predict potential human effects, from human studies (clinical or epidemiologic) or biobanked human samples, or the combination of all such approaches. The key to discovering or inferring biomarkers through computational means involves the identification or prediction of the molecular target(s) of the chemical(s) and the association of these targets with perturbed biological pathways. Two examples are given in this chapter: (1) inferring potential human biomarkers from animal toxicogenomics data, and (2) the identification of protein targets through computational means and associating these in one example with potential drug interactions and in another case with increasing the risk of developing certain human diseases.

Drug discovery of new anticancer therapeutics has shifted from a primary interest in cytotoxics to newer targeted therapies where specific molecular targets are the focus of both small-molecule and biotherapeutic modalities. The practice of front-loading toxicity testing has historically been used to accelerate programs; however, this must be tempered with the realization that many toxicities of targeted therapeutics relate to the actual mechanism of action for efficacy. Regulatory guidance is changing the requirements for toxicology testing for first-in-human clinical trials with the idea that translational assays may be discovered first in patients and these are then applied to longer-term toxicology studies. Whether mechanistic assays can be directly linked to toxicity in nonclinical studies has yet to be fully elucidated, but it seems clear that early screening for chemical-based toxicity will continue to be performed as part of the lead optimization process and that cardiotoxicity should be incorporated early on with targeted therapeutics because of the complex cross talk apparent in multiple signaling pathways. Keywords: targeted cancer therapeutics; toxicity screening in discovery; signaling pathways and toxicity; predicting human adverse events; first in human dose selection

A method is presented for homology modeling of proteins bearing weak sequence identity to proteins of known tertiary structure. To accommodate non-identical amino acids in the core region, the backbone of the structurally conserved core of the model protein is allowed to deviate from that of the template protein. We have expanded FOLDER, a distance geometry-based homology modeling method, to allow for such displacements in the structurally conserved core. Models are built by rigidly constraining the interatomic distances within a structurally conserved segment and by allowing the interatomic distances between these segments to vary by a "divergence factor". We test this method by simulating models of the beta-barrel domain D1 of CD4 and a four-helix bundle protein cytochrome b562 using the crystal structures of Bence-Jones protein and cytochrome c' as templates, respectively. In both cases, previously published structure-based sequence alignments were used for simulating models. The root-mean-square (r.m.s.) deviation of the backbone atoms in the common core between the templates and models was found to be a function of the imposed divergence factor. Our results demonstrate that this r.m.s. deviation results from the relative displacements of structurally conserved segments to accommodate the amino acid replacements in the core of the model protein. To test the integrity of the simulated structures we compared them with their respective crystal structures. The r.m.s. deviation of the backbone atoms in the core regions of the simulated models and their respective crystal structures is approximately 1.4 A. The r.m.s. deviation for all the backbone atoms in the models, including those in the structurally variable regions, which are modeled de novo, is 2.4 A for CD4 and 3.2 A for cytochrome b562 when compared with their respective X-ray structures.

We present an automated method incorporated into a software package, FOLDER, to fold a protein sequence on a given three-dimensional (3D) template. Starting with the sequence alignment of a family of homologous proteins, tertiary structures are modeled using the known 3D structure of one member of the family as a template. Homologous interatomic distances from the template are used as constraints. For nonhomologous regions in the model protein, the lower and the upper bounds for the interatomic distances are imposed by steric constraints and the globular dimensions of the template, respectively. Distance geometry is used to embed an ensemble of structures consistent with these distance bounds. Structures are selected from this ensemble based on minimal distance error criteria, after a penalty function optimization step. These structures are then refined using energy optimization methods. The method is tested by simulating the alpha-chain of horse hemoglobin using the alpha-chain of human hemoglobin as the template and by comparing the generated models with the crystal structure of the alpha-chain of horse hemoglobin. We also test the packing efficiency of this method by reconstructing the atomic positions of the interior side chains beyond C beta atoms of a protein domain from a known 3D structure. In both test cases, models retain the template constraints and any additionally imposed constraints while the packing of the interior residues is optimized with no short contacts or bond deformations. To demonstrate the use of this method in simulating structures of proteins with nonhomologous disulfides, we construct a model of murine interleukin (IL)-4 using the NMR structure of human IL-4 as the template. The resulting geometry of the nonhomologous disulfide in the model structure for murine IL-4 is consistent with standard disulfide geometry.

Traditional Chinese medicine (TCM) is a holistic approach to health that attempts to bring the body, mind and spirit into harmony. TCM is an essential part of the healthcare system in several Asian countries, and is considered a complementary or alternative medical system in most Western countries. An integration of the traditional Chinese and Western systems of medicine has begun in multiple medical centers internationally, and there is increasing evidence that several herbs and combinations of herbs used in TCM impart important pharmacological effects. The number of databases and compilations of herbs, herbal formulations, phytochemical constituents and molecular targets is increasing, primarily because of the widespread use of TCM in combination with Western drugs. The continued popularity of herbal remedies worldwide suggests that evidence-based research in this field, as well as information regarding the potential efficacy and safety of phytochemical constituents in herbs and TCM formulations, are essential, particularly when TCM is used in combination with other drugs. Herb-drug interactions are similar to drug-drug interactions in terms of their effects on ADME properties. Improvements in the knowledge of the molecular targets and metabolic pathways, as well as of the synergistic and inhibitory effects associated with important phytochemicals from herbs and herbal formulations, will lead to the development of rational approaches for the safe combination of healthcare systems from different cultures.

A mere 21 human phosphodiesterase (PDE) genes are responsible for modulating cellular levels of cAMP and cGMP in response to stimuli. Considering the importance of cAMP and cGMP to disparate physiological functions including visual response, smooth muscle relaxation, platelet aggregation, immune response, and cardiac contractibility, perhaps the 200 or more splice isoforms of PDE genes also play a major functional role. We profiled the human PDEs across 25 tissue samples using splice sensitive oligonucleotide microarrays containing probes for exons and exon–exon junctions. Our results suggest that PDEs exhibit tissue-specific differences in expression, as demonstrated by the high expression of PDE4B in skeletal muscle. At the splice variant level, the majority of PDE genes—notably 1A, 1C, 2A, 4C, 4D, 5A, 7A, 8A, 8B, 9A, 10A, and 11A—exhibited tissue-specific splicing with potential functional implications for PDE biology. This work validates expression of many EST transcripts, and confirms and expands on published findings based on PCR and cloning, illuminating some of the complexity of cAMP and cGMP processing.

We have catalogued the protein kinase complement of the human genome (the “kinome”) using public and proprietary genomic, complementary DNA, and expressed sequence tag (EST) sequences. This provides a starting point for comprehensive analysis of protein phosphorylation in normal and disease states, as well as a detailed view of the current state of human genome analysis through a focus on one large gene family. We identify 518 putative protein kinase genes, of which 71 have not previously been reported or described as kinases, and we extend or correct the protein sequences of 56 more kinases. New genes include members of well-studied families as well as previously unidentified families, some of which are conserved in model organisms. Classification and comparison with model organism kinomes identified orthologous groups and highlighted expansions specific to human and other lineages. We also identified 106 protein kinase pseudogenes. Chromosomal mapping revealed several small clusters of kinase genes and revealed that 244 kinases map to disease loci or cancer amplicons.

Summary: HyperTree is an application to visualize and navigate large trees in hyperbolic space. It includes color-coding, search mechanisms and navigational aids, as well as focus+context viewing, allowing enormous trees to fit within the fixed space of a computer screen or printed page. Availability: A demo is available online at http://www.kinase.com/tools/HyperTree.html Contact: sucha-sudarsanam@sugen.com

One of the most important questions in the protein folding problem is whether secondary structures are formed entirely by local interactions. One way to answer this question is to compare identical subsequences of proteins to see if they have identical structures. Such an exercise would also reveal a lower limit on the number of amino acids needed to form unique secondary structures. In this context, we have searched the April 1996 release of the Protein Data Bank for sequentially identical subsequences of proteins and compared their structures. We find that identical octamers can have different conformations. In addition, there are several examples of identical heptamers with different conformations, and the number of identical hexamers with different conformations has increased since the previous PDB releases. These observations imply that secondary structure can be formed entirely by non-local interactions and that an identical match of up to eight amino acids may not imply structural similarity. In addition to the larger context of the protein folding problem, these observations have implications for protein structure prediction methods.