Petricoin, Emanuel F.

Abstract Despite recent advances in cancer therapy, options for central nervous system metastasis (CM) remain limited and patient outcomes are dismal. The lack of tumor models that reflect the heterogeneity and complexity of human CM tumors is a barrier to understanding the underlying biology and to developing novel therapies. Here, we report the development and characterization of 39 patient derived xenograft (PDX) models of CM tumors with histological subtypes representing the disease spectrum. The PDX models successfully grew both in the flank and brains of mice, and retained their histology regardless of the site of implantation. Brain metastasis also formed in 5/12 models after intra-cardiac injection. Multi-omic analysis of the PDX tumor and the patient-matched CM tumor showed that PDXs largely retained the molecular profiles of the tumor from which they were derived, including the retention of aberrations in key driver genes and signaling pathways. PDXs differed primarily from the ungrafted patient tumors by a downregulation in immunity-related pathways, which was expected due to the use of an immuno-compromised host, arguing for the validity of using the flank as a site of implantation. Integrated analysis of mutations and gene expression changes in CM revealed an upregulation of neuronal signaling and DNA damage response, and downregulation of calcium signaling and inflammation. We also assessed the clonal evolution of the PDX from the original patient CM by tracing variant allele frequencies in copy number neutral regions. We found that the PDX tumors were heterogeneous and comprised of multiple tumor clones that were either unique or in common with the original tumor, and displayed evidence of clonal expansion. Two PDXs that underwent phylogeny assessment showed evidence of minor clone expansion. One such PDX was a tumor from a Her2+ breast cancer (CM13), which largely did not molecularly resemble its matched patient tumor but was confirmed to have some shared expressed variants. The PDX was nearly copy number neutral, which included loss of the Her2 amplicon, and had few mutations compared with its parent tumor. Most noteworthy, is that CM13 displayed unique molecular patterns, aggressive growth and was highly metastatic in vivo. From the flank, it led to multiple metastases, including micrometastasis in the brain. This suggests that minor clone takeover may confer tumor aggressiveness and potentiate brain metastasis. The large and diverse repertoire of PDXs developed and characterized in this study provides a new set of tools that will help deepen our understanding of CM and improve preclinical testing for CM therapies. Our data also show that expansion of minor clones could confer aggressive tumor behavior and may be important in tumor evolution in PDX models and human cancers. Citation Format: Ben Yi Tew, Christophe Legendre, Tim Triche, Gerald C. Gooden, Kyle N. Johnson, Rae Anne Martinez, Emanuel F. Petricoin, Mariaelena Pierobon, Joyce O'Shaughnessy, Cindy Osborne, Mark A. Schroeder, Daniel J. Ma, Mark Bernstein, Jann N. Sarkaria, Steven A. Toms, Bodour Salhia. Development and characterization of patient derived xenografts from central nervous system metastasis reveal minor clone expansion linked with aggressive tumor behavior [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 991.

Abstract Background: Although the efficacy of different immunotherapeutic agents varies greatly across tumor types, little is known about the role of different metastatic tumor microenvironments in modulating the expression of immune-checkpoints like PD-L1. Using metastatic breast cancer as a model, this work explored whether metastatic lesions derived from different patients, but colonizing the same host target organ, showed organ-specific immune signatures. Methods: Snap frozen metastatic lesions collected from 30 breast cancer patients enrolled in a prospective phase II multi-omic precision medicine clinical trial (“Side-Out II”) were used for this analysis. Sites of metastasis were: liver (LI) (n=10), skin/chest wall (SC) (n=10), and other lesions (OL) (n=10) [including lymph nodes (LN) (n=4), lung (LU) (n=3), and intra-abdominal lesions (IA) (n=3)]. Pure tumor epithelia were isolated from each specimen via Laser Capture Microdissection (LCM). PD-L1 expression (Antibody Clone E1L3N) was quantitatively measured using the Reverse Phase Protein Microarray (RPPA), a high-throughput, multiplex immunoassay that provides operator-independent quantitative data starting from a small amount of biological material. Proportion of patients with PD-L1 expression above the median and the 75th percentile of the population was compared across metastatic sites. Results: PD-L1 expression had a >5 fold dynamic range across breast cancer-derived metastatic lesions with different distribution across metastatic sites. Specifically, PD-L1 expression of 6/10 LIs, 5/10 OLs (including 2 LU and 3 LN positive lesions), and 4/10 SC metastases was greater than the population median of all metastases combined. LI metastases were the most represented group above the 75th percentile of the population (4/10), followed by OS (2/10) (including 1 LU and 1 LN positive lesions), and SC (1/10). Of the 15 lesions above the population median, 6 were LI (40%), 5 were OS (33%), and 4 were SC (27%); of the 7 metastases above the 75th percentile, 4 were LI (57%), 2 were OS (29%), and 1 was SC (14%). Conclusions: The LCM-RPPA workflow can capture PD-L1 protein expression on a continuous quantitative scale and provide a broader understanding of the dynamic range of expression. Our data suggest that organ specific microenvironments in which metastatic lesions develop may strongly influence overall PD-L1 tumor cell expression. If different metastatic host organs can modulate PD-L1 expression, location of the metastatic lesion should be kept in consideration when selecting patients that are candidates for immunotherapy. Such hypotheses should be further validated in prospective clinical trial where quantitative PD-L1 expression of different metastatic lesions is evaluated along with response to targeted immuno-oncology compounds. Citation Format: Mariaelena Pierobon, K Alex Hodge, Elisa Baldelli, Steven Anthony, Nicholas J. Robert, Donald W. Northfelt, Mohammad J. Jahanzeb, Linda Vocila, Lance A. Liotta, Bryant Dunetz, Emanuel F. Petricoin. PD-L1 expression differs across cancer metastatic sites from breast tumors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 3129.

Abstract Purpose: Little is known about the molecular signatures associated with specific metastatic sites in breast cancer. Using comprehensive multi-omic molecular profiling, we assessed whether alterations or activation of the PI3K–AKT–mTOR pathway is associated with specific sites of breast cancer metastasis. Experimental Design: Next-generation sequencing–based whole-exome sequencing was coupled with reverse-phase protein microarray (RPPA) functional signaling network analysis to explore the PI3K–AKT–mTOR axis in 32 pretreated breast cancer metastases. RPPA-based signaling data were further validated in an independent cohort of 154 metastatic lesions from breast cancer and 101 unmatched primary breast tumors. The proportion of cases with PI3K–AKT–mTOR genomic alterations or signaling network activation were compared between hepatic and nonhepatic lesions. Results: PIK3CA mutation and activation of AKT (S473) and p70S6K (T389) were detected more frequently among liver metastases than nonhepatic lesions (P < 0.01, P = 0.056, and P = 0.053, respectively). However, PIK3CA mutations alone were insufficient in predicting protein activation (P = 0.32 and P = 0.19 for activated AKT and p70S6K, respectively). RPPA analysis of an independent cohort of 154 tumors confirmed the relationship between pathway activation and hepatic metastasis [AKT (S473), mTOR (S2448), and 4EBP1 (S65); P < 0.01, P = 0.02, and P = 0.01, respectively]. Similar results were also seen between liver metastases and primary breast tumors [AKT (S473) P < 0.01, mTOR (S2448) P < 0.01, 4EBP1 (S65) P = 0.01]. This signature was lost when primary tumors were compared with all metastatic sites combined. Conclusions: Breast cancer patients with liver metastasis may represent a molecularly homogenized cohort with increased incidence of PIK3CA mutations and activation of the PI3K–AKT–mTOR signaling network. Clin Cancer Res; 23(16); 4919–28. ©2017 AACR.

Abstract Background: We hypothesize that response to the pan-ERBB inhibitor, neratinib (N), may be predicted by pre-treatment HER2-EGFR signaling. In the I-SPY 2 TRIAL, N graduated in the HR-/HER2+ signature. All patients received at least standard chemotherapy. For HER2+ patients, N was administered in place of trastuzumab. We evaluated 18 HER family signaling proteins as biomarkers of N response using reverse phase protein microarray (RPMA) data from pre-treatment LCM purified tumor epithelium. Methods: 168 patients (N: 106, concurrent controls: 62) had RPMA and pCR data. 18 biomarkers relating to HER family signaling were evaluated: AKT S473, AKT T308, EGFR, EGFR Y1068, EGFR Y1148, EGFR Y1173, EGFR Y992, ERBB2, ERBB2 Y1248, ERBB3 total, ERBB3 Y1289, ERK1/2 T202/Y204, Heregulin, mTOR, mTOR S2448, PI3K p85 Y458/p55 Y199, PTEN S380, and SHC Y317. We assessed association between biomarker and response in the N and control arms alone (likelihood ratio test), and relative performance between arms (biomarker x treatment interaction) using a logistic model. Analysis was also performed adjusting for HR/HER2 status. In an exploratory analysis, we selected the marker with the greatest interaction (phosphorylated EGFR (Y1173)) to dichotomize patients optimally based on the data and assessed it in the context of the graduating signature by adding the EGFR Y1173-High patients to the HR-/HER2+ subtype and evaluating the treatment effect in this ‘biomarker-positive’ group. Our study is exploratory with no claims for generalizability of the data and does not account for multiplicities. Statistical calculations are descriptive (e.g. p-values are measures of distance with no inferential content). Results: 7 HER pathway markers (EGFR Y1068, EGFR Y1173, EGFR Y992, ERBB2 total, ERBB2 Y1248, ERBB3 Y1289, SHC Y317) are associated with response in the N but not the control arm. However, the difference in performance between arms did not reach significance by permutation testing. Adjusting for HR/HER2 status, EGFR Y1173 shows a significant biomarker x treatment interaction (p = 0.049). In an exploratory analysis, we dichotomized patients by their EGFR Y1173 levels and evaluated the distribution of pCR rates (Table 1). Neratinib (n=106)Control (n=62)EGFR Y1173 Low (n=31)EGFR Y1173 High (n=75)EGFR Y1173 Low (n=29)EGFR Y1173 High (n=33)HR-/HER2+ (n=28)0 / 412 / 181 / 11 / 5Not HR-/HER2+ (n=140)3 / 2724 / 575 / 285 / 28Table 1 OR between EGFR Y1173 groups in the N relative to control arm is 10.1. When EGFR Y1173 High patients are added to the graduating HR-/HER2+ subset, the OR associated with treatment is 3.2 and is comparable to that in the HR-/HER2+ signature (OR: 2.1), while increasing the prevalence of biomarker-positive patients by ∼50%. Evaluation of EGFR Y1173 under the I-SPY 2 Bayesian model is pending. Conclusion: Our sample size is too small to draw definitive conclusions. Our exploratory analysis reveals that HER family phosphoproteins associate with response to N, but only phosphorylated EGFR Y1173 appears to add value to the graduating signature. Given that this biomarker would expand the patient population that may benefit, it merits evaluation in other ongoing trials of neratinib. Citation Format: Julia D Wulfkuhle, Christina Yau, Denise M Wolf, Rosa I Gallagher, Ashish Sanil, Lamorna Brown-Swigart, Susan Flynn, Gillian Hirst, I-SPY 2 TRIAL Investigators, Meredith Buxton, Angela DeMichele, Nola Hylton, William F Symmans, Laura van't Veer, Douglas Yee, Melissa Paoloni, Laura Esserman, Donald Berry, Minetta C Liu, John W Park, Emanuel F Petricoin III. Evaluation of HER family protein signaling network as a predictive biomarker for pCR for breast cancer patients treated with neratinib in the I-SPY 2 trial [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr P3-06-15.

Abstract Background: The development of distant metastases is the strongest prognostic factor associated with cancer mortality. Customization of treatment based on molecular profiling of a patient’s primary tumor has yielded promising results and has opened a new paradigm for treating patients with advanced disease. Nevertheless profiling the primary tumor may not reflect the actionable molecular drivers of the metastatic lesions. In the last few years a large number of genomic and proteomic studies have demonstrated that at the molecular level metastatic lesions significantly differ from matched primary tumors. The impact of metastasis site organ microenvironments on breast cancer primary tumors has been only partially explored. Since most targeted therapies work by modulating aberrantly-activated protein kinase signaling, the aim of this study was to utilize reverse phase protein arrays (RPPA) to explore whether metastatic lesions derived from different patients, but invading the same host target organ showed similarities in their signaling architecture and presented with organ-specific targetable signatures. Methods: Snap frozen material collected from 14 metastatic breast cancer patients enrolled in a prospective phase II trial ("Side Out II") were used for this analysis. Sites of metastasis were: liver (n=7), skin/chest wall (n=4), and lung (n=3). All samples were subjected to Laser Capture Microdissection (LCM) and RPPA to measure the activation/phosphorylation levels of 12 FDA approved drug targets and linked substrates. Results: Among the 7 patients with liver metastases, p70S6K, c-Abl, ERK 1/2 were highly phosphorylated in 5 cases (71.4%), ErbB2/HER2 in 4 cases (57.1%), and AKT in 3 cases (42.8%). On the contrary, of the 4 patients with skin/chest wall lesions 3 showed high activation of ErbB2/HER2, ErbB3/HER3, and ERK 1/2 (75.0%) and 2 of EGFR (50.0%). Of note, none of the skin/chest wall lesions showed activation of AKT and only 1 case presented with activated p70S6K (25.0%). Lung metastasis showed an overall low activation of all 12 analytes measured. Discussion: Although this data is based on a relatively small number of samples and needs further validation, our initial analysis revealed potentially important trends that could have an impact in the prioritization and selection of targeted agents for metastatic breast cancer trials. We found that the distribution of the 12 FDA approved drug targets and downstream substrates varies between different metastatic sites. Thus, the emergent hypothesis from this work is that breast cancer patients with liver metastases may be more susceptible to AKT-mTOR directed targeted therapeutics and patients with skin/chest wall metastasis may benefit from therapies that target EGFR and/or downstream ERK signaling. Citation Format: Mariaelena Pierobon, Stephen P Anthony, K Alex Reeder, Nicholas J Robert, Donald W Northfelt, Mohammad Jahanzeb, Linda Vocila, Julia D Wulfkuhle, Bryant Dunetz, Lance A Liotta, Emanuel F Petricoin. Pathway activation mapping of metastatic breast cancer identifies potential organ-specific signatures: Implications for patient stratification to targeted treatment [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr P1-07-23.

AbstractThe many improvements in breast cancer therapy in recent years have so lowered rates of recurrence that it is now difficult or impossible to conduct adequately powered adjuvant clinical trials. Given the many new drugs and potential synergistic combinations, the neoadjuvant approach has been used to test benefit of drug combinations in clinical trials of primary breast cancer. A recent FDA-led meta-analysis showed that pathologic complete response (pCR) predicts disease-free survival (DFS) within patients who have specific breast cancer subtypes. This meta-analysis motivated the FDA's draft guidance for using pCR as a surrogate endpoint in accelerated drug approval. Using pCR as a registration endpoint was challenged at ASCO 2014 Annual Meeting with the presentation of ALTTO, an adjuvant trial in HER2-positive breast cancer that showed a nonsignificant reduction in DFS hazard rate for adding lapatinib, a HER-family tyrosine kinase inhibitor, to trastuzumab and chemotherapy. This conclusion seemed to be inconsistent with the results of NeoALTTO, a neoadjuvant trial that found a statistical improvement in pCR rate for the identical lapatinib-containing regimen. We address differences in the two trials that may account for discordant conclusions. However, we use the FDA meta-analysis to show that there is no discordance at all between the observed pCR difference in NeoALTTO and the observed HR in ALTTO. This underscores the importance of appropriately modeling the two endpoints when designing clinical trials. The I-SPY 2/3 neoadjuvant trials exemplify this approach. Clin Cancer Res; 21(13); 2911–5. ©2015 AACR.