Kathryn A. Kolquist

Clinical validation of a gene expression signature that differentiates benign nevi from malignant melanoma

Histopathologic examination is sometimes inadequate for accurate and reproducible diagnosis of certain melanocytic neoplasms. As a result, more sophisticated and objective methods have been sought. The goal of this study was to identify a gene expression signature that reliably differentiated benign and malignant melanocytic lesions and evaluate its potential clinical applicability. Herein, we describe the development of a gene expression signature and its clinical validation using multiple independent cohorts of melanocytic lesions representing a broad spectrum of histopathologic subtypes.

An independent validation of a gene expression signature to differentiate malignant melanoma from benign melanocytic nevi

Recently, a 23‐gene signature was developed to produce a melanoma diagnostic score capable of differentiating malignant and benign melanocytic lesions. The primary objective of this study was to independently assess the ability of the gene signature to differentiate melanoma from benign nevi in clinically relevant lesions.

Analytical validation of a melanoma diagnostic gene signature using formalin- fixed paraffin-embedded melanocytic lesions

AIM These studies were to validate the analytical performance of a gene expression signature that differentiates melanoma and nevi, using RNA expression from 14 signature genes and nine normalization genes that generates a melanoma diagnostic score (MDS). MATERIALS & METHODS Formalin-fixed paraffin-embedded melanocytic lesions were evaluated in these studies. RESULTS The overall SD of the assay was determined to be 0.69 MDS units. Individual amplicons within the signature had an average amplification efficiency of 92% and a SD less than 0.5 CT. The MDS was reproducible across a 2000-fold dilution range of input RNA. Melanin, an inhibitor of PCR, does not interfere with the signature. CONCLUSION These studies indicate this signature is robust and reproducible and is analytically validated on formalin-fixed paraffin-embedded melanocytic lesions.

Cell cycle progression score is a marker for five-year lung cancer-specific mortality risk in patients with resected stage I lung adenocarcinoma

Purpose The goals of our study were (a) to validate a molecular expression signature (cell cycle progression [CCP] score and molecular prognostic score [mPS; combination of CCP and pathological stage {IA or IB}]) that identifies stage I lung adenocarcinoma (ADC) patients with a higher risk of cancer-specific death following curative-intent surgical resection, and (b) to determine whether mPS stratifies prognosis within stage I lung ADC histological subtypes. Methods Formalin-fixed, paraffin-embedded stage I lung ADC tumor samples from 1200 patients were analyzed for 31 proliferation genes by quantitative RT-PCR. Prognostic discrimination of CCP score and mPS was assessed by Cox proportional hazards regression, using 5-year lung cancer–specific mortality as the primary outcome. Results In multivariable analysis, CCP score was a prognostic marker for 5-year lung cancer–specific mortality (HR=1.6 per interquartile range; 95% CI, 1.14–2.24; P=0.006). In a multivariable model that included mPS instead of CCP, mPS was a significant prognostic marker for 5-year lung cancer–specific mortality (HR=1.77; 95% CI, 1.18–2.66; P=0.006). Five-year lung cancer–specific survival differed between low-risk and high-risk mPS groups (96% vs 81%; P<0.001). In patients with intermediate-grade lung ADC of acinar and papillary subtypes, high mPS was associated with worse 5-year lung cancer–specific survival (P<0.001 and 0.015, respectively), compared with low mPS. Conclusion This study validates CCP score and mPS as independent prognostic markers for lung cancer–specific mortality and provides quantitative risk assessment, independent of known high-risk features, for stage I lung ADC patients treated with surgery alone.

The influence of a gene expression signature on the diagnosis and recommended treatment of melanocytic tumors by dermatopathologists

AbstractIt is well documented that histopathologic examination is sometimes inadequate for accurate and reproducible diagnosis of certain melanocytic neoplasms. Recently, a 23-gene expression signature has been clinically validated as an adjunctive diagnostic test to differentiate benign nevi from malignant melanomas. This study aimed to quantify the impact of this test on diagnosis and treatment recommendations made by dermatopathologists.Diagnostically challenging melanocytic lesions encountered during routine dermatopathology practice were submitted for gene expression testing and received a melanoma diagnostic score (MDS). Submitting dermatopathologists completed a survey documenting pre-test diagnosis, level of diagnostic confidence, and recommendations for treatment. The survey was repeated after receiving the MDS. Changes between the pre- and post-test surveys were analyzed retrospectively.When the MDS was available as part of a comprehensive case evaluation in diagnostically challenging cases, definitive diagnoses were increased by 56.6% for cases that were initially indeterminate and changes in treatment recommendations occurred in 49.1% of cases. Treatment recommendations were changed to align with the test result in 76.6% of diagnostically challenging cases.The MDS impacts diagnosis and treatment recommendations by dermatopathologists confronted with diagnostically challenging melanocytic lesions. Increased data are needed in order to completely understand how use of the MDS will translate from dermatopathology to clinical practice.

Analytical Validation of a Cell Cycle Progression Signature Used as a Prognostic Marker in Prostate Cancer

Background: Prostate cancer is the most common cancer in men in the developed world. Appropriate clinical care requires accurate prognostic information to determine whether definitive treatment or conservative management is most appropriate for a given patient. We previously demonstrated that a gene expression signature, which measures the RNA expression of 31 cell cycle progression (CCP) genes and generates a CCP score, is a robust predictor of patient outcome in cohorts of conservatively managed patients diagnosed by needle biopsy or transurethral resection of the prostate. Methods: These current studies represent the analytical validation of this gene signature, for the testing of either formalin-fixed paraffin-embedded (FFPE) prostate resection tissue (radical prostatectomy, RP) or FFPE prostate needle biopsy samples. Results: The measured standard deviation (SD) of the signature was determined to be 0.1 score units, representing 1.6% of the range of scores observed within previous clinical validation studies. Individual amplicons for all genes within the signature had a SD <1 CT, with a median SD of 0.52 CT’s. We observed the median amplification efficiency for all genes was 92.6%. The linear range of the signature was over a ~260-fold range of RNA concentrations. We observed that 100% of RP samples and 99.8% of needle biopsy samples produced sufficient RNA for testing, when RNA was extracted from 7,525 recent prostate samples. Finally, RNA samples were able to reproduce similar CCP scores when stored for up to 8 weeks. Conclusion: These studies indicate that this prognostic gene signature is robust and reproducible, and is analytically validated for use on FFPE prostate biopsy radical prostatectomy samples.

Analytical validation of a proliferation-based molecular signature used as a prognostic marker in early stage lung adenocarcinoma

AIMS The aim of these studies was to validate the analytical performance of a cell cycle progression (CCP) gene signature that provides prognostic information for early stage lung adenocarcinomas. MATERIALS & METHODS Formalin-fixed paraffin-embedded (FFPE) lung resections were evaluated by quantitative RT-PCR for the expression of 31 target and 15 housekeeper genes comprising the CCP score. RESULTS The signature had a standard deviation (SD) of 0.06 score units and a dynamic range spanning CCP scores between -13 and 14. The average amplicon efficiencies for target and housekeeper genes were 107% and 105%, respectively. All but one amplicon had a SD <0.5 CT. CONCLUSION These studies demonstrate that the gene signature is robust and reproducible, making it suitable for use in a clinical setting.

Abstract C53: Homologous recombination deficiency (HRD) of high grade serous ovarian tumors from the NOVA Phase III clinical study

Background: Genome-wide analysis was conducted on tumors obtained from patients enrolled in the NOVA study, a Phase 3 clinical trial evaluating the PARP inhibitor niraparib as a maintenance treatment in patients with platinum-sensitive ovarian cancer. Homologous recombination deficiency (HRD) and mutations in DNA damage repair genes were evaluated. Material and methods: DNA was extracted from formalin-fixed paraffin-embedded (FFPE) tumor tissue and used to create libraries that were hybridized to a custom Agilent SureSelect capture array carrying probes for 54,091 single nucleotide polymorphism sites distributed across the human genome, as well as probes targeting 43 genes involved in DNA repair, including BRCA1 and BRCA2. The captured and enriched DNA was sequenced on an Illumina HiSeq 2500 sequencer. Sequences covering SNP positions were used to generate allelic imbalance profiles. Measures of genomic instability, including determination of an HRD score (integer value of 0-100), were calculated using allelic imbalance profiles and determination of loss of heterozygosity (LOH) by allele-specific copy number (ASCN). A previously identified HRD threshold score of 42 was used to define HRD positivity in the absence of a BRCA mutation. Results: The NOVA study is a Phase 3, multicenter, randomized, double-blind, placebo-controlled study of niraparib as maintenance therapy in ovarian cancer patients who have either gBRCAmut or a tumor with high-grade serous histology and who have responded to their most recent chemotherapy containing a platinum agent. Tumor BRCA mutational status, HRD score and genomic sequencing of 43 DNA repair genes were obtained from tumor samples from both gBRCAmut and non-gBRCAmut cohorts. In the gBRCAmut cohort, HRD analysis of the tumor confirmed the presence of a deleterious or suspected deleterious mutation in all cases. In addition, an HRD score ≥ 42 and the presence of a deleterious mutation in TP53 with loss of heterozygosity (LOH) were observed in nearly all tumors. In the non-gBRCA cohort, somatic BRCA mutations were observed in approximately 13% of tumors, and approximately half of tumors with no evidence of a BRCA mutation had a high HRD score. In both cohorts, the use of three scoring algorithms (LOH, telomeric allelic imbalance [TAI], large-scale state transitions [LST]), was more predictive of BRCA mutational status than LOH alone. Additional genomic sequencing identified deleterious mutations with LOH in DNA repair genes, such as BRIP1, CDK12, RAD51C, PTEN, and RAD51D, with many tumors exhibiting multiple deleterious mutations. Conclusions: High grade serous ovarian cancer is characterized by a high degree of genomic instability. Genomic analysis in the clinical setting is able to identify patients with both germline and somatic BRCA mutations, in addition to BRCAwt tumors with other genetic defects that may be sensitive to agents exploiting deficiencies in HR. Citation Format: Keith Wilcoxen, Christopher Neff, Victor Abkevich, Joshua Timothy Jones, Kathryn Kolquist, Michael Mirza, Jerry Lanchbury, Keith Mikule, Shefali Agarwal, Anne-Renee Hartman, Alexander Gutin, Kirsten Timms. Homologous recombination deficiency (HRD) of high grade serous ovarian tumors from the NOVA Phase III clinical study. [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2015 Nov 5-9; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2015;14(12 Suppl 2):Abstract nr C53.