Terry Hyslop

<jats:title>Abstract</jats:title> <jats:p>Cancer proteogenomics combines genomics, transcriptomics and mass spectrometry-based proteomics to gain insights into cancer biology and treatment responsiveness. While proteogenomics analyses have already shown great potential to deepen our understanding of cancer tissue complexity and signaling, how a patient’s tumor changes upon treatment has largely been the province of genomics. This is due to technical difficulties associated with doing proteogenomic analysis on clinic-derived core-needle biopsies. To address this critical need, we have developed a “microscaled” proteogenomics approach for tumor-rich OCT-embedded core needle biopsies. Tissue-sparing specimen processing (“Biopsy Trifecta EXTraction”, BioTExt) and microscaled proteomics (MiProt) methodologies allowed generation of deep-scale proteogenomics datasets, with copy number and transcript information for &amp;gt;20,000 genes and mass spectrometry-based identification and quantification of nearly all expressed proteins in a tumor (&amp;gt;10,000 proteins) and more than &amp;gt;20,000 phosphosites starting with just 25 micrograms of protein per sample. In order to understand the capabilities and limitations our our approach relative to more conventional deepscale proteomics requiring &amp;gt;10X more starting material, we compared preclinical patient derived xenograft (PDX) models at conventional scale with data obtained by core-needle biopsy of the same tissues. Comparable depth and biological insights were obtained from the cores relative to surgically resected tumors. As a proof-of-concept for implementation in clinical trials, we applied microscaled proteogenomic methods to a small-scale clinical study where biopsies were accrued from patients with ERBB2+ locally advanced breast cancer before and 48 to 72 hours after the first dose of neoadjuvant Trastuzumab-based chemotherapy. Multi-omics comparisons were conducted between samples associated with residual disease versus samples associated with complete pathological response. Integrative, microscaled proteogenomic analyses efficiently diagnosed the molecular bases of diverse candidate treatment resistance mechanisms including: 1) absence of ERBB2 amplification (false-ERBB2+); 2) insufficient ERBB2 activity for therapeutic sensitivity despite ERBB2 amplification (pseudo-ERBB2+); 3) resistance features in true-ERBB2+ cases including androgen receptor signaling, mucin expression and an inactive immune microenvironment; 4) lack of acute phospho-ERBB2 down-regulation in non-pCR cases. In summary, we have developed a robust proteogenomics pipeline well suited for large-scale cancer clinical studies to identify potential resistance mechanism in patients. We conclude that microscaled cancer proteogenomics could improve diagnostic precision in the clinical setting.</jats:p> <jats:p>Citation Format: Shankha Satpathy, Eric Jaehnig, Krug Karsten, Beom-Jun Kim, Alexander Saltzman, Doug Chan, Kimberly Holloway, Meenakshi Anurag, Chen Huang, Purba Singh, Ari Gao, Noel Namai, Yongchao Dou, Bo Wen, Suhas Vasaikar, David Mutch, Mark Watson, Cynthia Ma, Foluso Ademuyiwa, Mothaffar Rimawi, Jeremy Hoog, Samuel Jacobs, Anna Malovannaya, Terry Hyslop, D.R Mani, Charles Perou, George Miles, Bing Zhang, Michael Gillette, Steven Carr, Matthew Ellis. Microscaled proteogenomic methods for precision oncology [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr GS2-05.</jats:p>

BACKGROUND: We examined the prognostic significance of circulating tumor cell (CTC) dynamics during treatment in metastatic breast cancer (MBC) patients receiving first-line chemotherapy. METHODS: Serial CTC data from 469 patients (2202 samples) were used to build a novel latent mixture model to identify groups with similar CTC trajectory (tCTC) patterns during the course of treatment. Cox regression was used to estimate hazard ratios for progression-free survival (PFS) and overall survival (OS) in groups based on baseline CTCs, combined CTC status at baseline to the end of cycle 1, and tCTC. Akaike information criterion was used to select the model that best predicted PFS and OS. RESULTS: Latent mixture modeling revealed 4 distinct tCTC patterns: undetectable CTCs (56.9% ), low (23.7%), intermediate (14.5%), or high (4.9%). Patients with low, intermediate, and high tCTC patterns had statistically significant inferior PFS and OS compared with those with undetectable CTCs (P < .001). Akaike Information Criterion indicated that the tCTC model best predicted PFS and OS compared with baseline CTCs and combined CTC status at baseline to the end of cycle 1 models. Validation studies in an independent cohort of 1856 MBC patients confirmed these findings. Further validation using only a single pretreatment CTC measurement confirmed prognostic performance of the tCTC model. CONCLUSIONS: We identified 4 novel prognostic groups in MBC based on similarities in tCTC patterns during chemotherapy. Prognostic groups included patients with very poor outcome (intermediate + high CTCs, 19.4%) who could benefit from more effective treatment. Our novel prognostic classification approach may be used for fine-tuning of CTC-based risk stratification strategies to guide future prospective clinical trials in MBC.

<jats:title>Abstract</jats:title> <jats:p>Background: Triple-negative breast cancer is an aggressive disease with limited treatment options beyond chemotherapy. In CALGB 40603, adding either carboplatin (Cb) or bevacizumab (Bev) to standard neoadjuvant chemotherapy (NACT) increased pathologic complete response (pCR) rates in TNBC and the subset of Basal-like breast cancers. Transcriptomic studies have previously found that evidence of immune activation was significantly associated with pCR. We are now examining frequency and pCR impact of germline and somatic mutations as secondary analyses.</jats:p> <jats:p>Methods: 281 pretreatment tumors and matched blood samples were sequenced using a 1,123 cancer-associated gene panel. Germline and somatic mutations were determined and somatic mutations were further evaluated and filtered using a previously identified list of likely false positives (Bailey, Cell 2019) and RNA expression of the variant. We examined mutation frequency and association of mutations with overall in-breast pCR (result available for 274 of our cohort) and for benefit of addition of Cb or Bev on pCR.</jats:p> <jats:p>Results: The pCR rate in patients with DNA sequencing was 53.6% in TNBC (147 pCR, 127 non-pCR) and 53.3% in Basal-like (131 pCR, 115 non-pCR). Total number of somatic mutations per sample from our targeted cancer panel ranged from one to 1,041, with a median of 18 mutations per patient. Ten samples had more than 100 mutations, of which six had POLE or POLD1 mutations. Focusing on non-silent coding mutations reduced the number of variants in the entire cohort from 10,483 to 4,275. A second filter was used to remove likely false positives. Mutations in TP53 were identified in 90% of samples. Only two other genes - KMT2C (12%) and MACF1 (11%) - were mutated in more than 10% of the samples. Mutations were also identified in NF1 (8%), PIK3CA (8%), PTEN (7%), FAT1 (7%), and RB1 (7%). Somatic alterations in BRCA1 and BRCA2 were observed in 6% and 5% of samples, respectfully. We also found that 40% of the tumors had one or more mutations in a set of 18 histone modification genes. However, these mutations were not associated with pCR in the entire cohort or in the Basal-like subset (n=246). The mutations were also not associated with response to Cb or Bev. The germline was also analyzed for inherited variants. We identified 35 (12%) patients with pathogenic or likely pathogenic variants in BRCA1 (n=23), BRCA2 (n=6), PALB2 (n=3), ATM (n=2), and CHEK2 (n=1). Germline variants were not associated with pCR in the entire cohort (n=33) or Basal subset (n=32), but there was a trend toward higher pCR in germline carriers (n=15) versus non-germline carriers (n=123) within the Bev-treated cohort (80% vs 54%, p=0.09). The pCR rate in patients with a pathogenic germline or somatic BRCA1/2 mutation (n=55) was similar to that for the overall study population (53% pCR rate), as was the increase in pCR seen with the addition of Cb (61% vs. 41% pCR rate) (Sikov, JCO 2015).</jats:p> <jats:p>Conclusions: Analysis of recurrently mutated genes in TNBC among patients treated in CALGB 40603 revealed high rates of mutation and very high TP53 mutation frequency but failed to identify mutations significantly associated with pCR rate. Integrated analyses with copy number, immune and other RNA features are underway. Support for the project came from U10CA180821, U10CA180882; BCRF, Susan G. Komen, and Genentech. https://acknowledgments.alliancefound.org</jats:p> <jats:p>ClinicalTrials.gov Identifier: NCT00861705</jats:p> <jats:p>Citation Format: Katherine A Hoadley, Bradford C. Powell, Dona Kanavy, David Marron, Lisle E. Mose, Terry Hyslop, Donald A. Berry, Olwen Hahn, Sara M. Tolaney, William M. Sikov, Charles M. Perou, Lisa A. Carey. Mutational analysis of triple-negative breast cancer (TNBC): CALGB 40603 (Alliance) [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P4-05-03.</jats:p>

BACKGROUND. The incidence of ductal carcinoma in situ (DCIS) has steadily increased, as have concerns regarding overtreatment. Active surveillance is a novel treatment strategy that avoids surgical excision, but identifying patients with occult invasive disease who should be excluded from active surveillance is challenging. Radiologists are not typically expected to predict the upstaging of DCIS to invasive disease, though they might be trained to perform this task. OBJECTIVE. The purpose of this study was to determine whether a mixed-methods two-stage observer study can improve radiologists' ability to predict upstaging of DCIS to invasive disease on mammography. METHODS. All cases of DCIS calcifications that underwent stereotactic biopsy between 2010 and 2015 were identified. Two cohorts were randomly generated, each containing 150 cases (120 pure DCIS cases and 30 DCIS cases upstaged to invasive disease at surgery). Nine breast radiologists reviewed the mammograms in the first cohort in a blinded fashion and scored the probability of upstaging to invasive disease. The radiologists then reviewed the cases and results collectively in a focus group to develop consensus criteria that could improve their ability to predict upstaging. The radiologists reviewed the mammograms from the second cohort in a blinded fashion and again scored the probability of upstaging. Statistical analysis compared the performances between rounds 1 and 2. RESULTS. The mean AUC for reader performance in predicting upstaging in round 1 was 0.623 (range, 0.514-0.684). In the focus group, radiologists agreed that upstaging was better predicted when an associated mass, asymmetry, or architectural distortion was present; when densely packed calcifications extended over a larger area; and when the most suspicious features were focused on rather than the most common features. Additionally, radiologists agreed that BI-RADS descriptors do not adequately characterize risk of invasion, and that microinvasive disease and smaller areas of DCIS will have poor prediction estimates. Reader performance significantly improved in round 2 (mean AUC, 0.765; range, 0.617-0.852; p = .045). CONCLUSION. A mixed-methods two-stage observer study identified factors that helped radiologists significantly improve their ability to predict upstaging of DCIS to invasive disease. CLINICAL IMPACT. Breast radiologists can be trained to better predict upstaging of DCIS to invasive disease, which may facilitate discussions with patients and referring providers.

BACKGROUND: Implementation methods of risk-stratified cancer screening guidance throughout a health care system remains understudied. OBJECTIVE: Conduct a preliminary analysis of the implementation of a risk-stratified prostate cancer screening algorithm in a single health care system. DESIGN: Comparison of men seen pre-implementation (2/1/2016-2/1/2017) vs. post-implementation (2/2/2017-2/21/2018). PARTICIPANTS: Men, aged 40-75 years, without a history of prostate cancer, who were seen by a primary care provider. INTERVENTIONS: The algorithm was integrated into two components in the electronic health record (EHR): in Health Maintenance as a personalized screening reminder and in tailored messages to providers that accompanied prostate-specific antigen (PSA) results. MAIN MEASURES: Primary outcomes: percent of men who met screening algorithm criteria; percent of men with a PSA result. Logistic repeated measures mixed models were used to test for differences in the proportion of individuals that met screening criteria in the pre- and post-implementation periods with age, race, family history, and PSA level included as covariates. KEY RESULTS: During the pre- and post-implementation periods, 49,053 and 49,980 men, respectively, were seen across 26 clinics (20.6% African American). The proportion of men who met screening algorithm criteria increased from 49.3% (pre-implementation) to 68.0% (post-implementation) (p < 0.001); this increase was observed across all races, age groups, and primary care clinics. Importantly, the percent of men who had a PSA did not change: 55.3% pre-implementation, 55.0% post-implementation. The adjusted odds of meeting algorithm-based screening was 6.5-times higher in the post-implementation period than in the pre-implementation period (95% confidence interval, 5.97 to 7.05). CONCLUSIONS: In this preliminary analysis, following implementation of an EHR-based algorithm, we observed a rapid change in practice with an increase in screening in higher-risk groups balanced with a decrease in screening in low-risk groups. Future efforts will evaluate costs and downstream outcomes of this strategy.