Acquired resistance

EGFR-mutant lung cancers (LC) serve as the prototype oncogene-driven cancer and mechanisms of resistance to EGFR therapy are well-established. Selective pressure of EGFR inhibitors can induce lineage plasticity to escape oncogene-dependence. While we have shown EGFR/TP53/RB1-mutations are necessary but insufficient to induce small cell transformation, no other genomic signatures and transcriptomic/epigenomic effectors that induce transformation have been identified.

Using parallel exome sequencing, bulk RNA sequencing, and bisulfite sequencing, we will characterize the biomarkers of transformation in EGFR/TP53/RB1-mutant LC and mixed adenocarcinoma/squamous cell tumors. We will also utilize single-cell RNA sequencing to identify the transcriptional changes associated with transformation and to identify distinct cell types within a heterogenous tumor. We will reconstruct a phenotypic landscape to identify subclones in a transitional state between histologies. The information obtained from these studies will nominate novel biomarkers of lineage plasticity that will be validated and proposed for study in clinical trials.

Despite high tumor response rates, patients treated with osimertinib inevitably develop disease progression. Mechanisms of both intrinsic, adaptive, and acquired drug resistance remain incompletely understood on a genomic and proteomic level. Our overall objective is to find new targeted treatments and drug combinations that can tackle cancer evolution and drug resistance.

We seek to:

1. Understand biological determinants of response prior to therapy (poor prognosis features)
2. Utilize correct combination therapies in the first line to enhance response, based on biological features of subpopulations present at the start
3. Reprogram cells toward a more drug-sensitive state through understanding network plasticity.

More specifically, our aims are:

1. To define the genomic and proteomic characteristics of drug-tolerant persister cells (DTPCs) under dynamic treatment with osimertinib. Our goal is to identify new vulnerabilities in EGFR-mutant tumors which, when therapeutically targeted, can maximize the depth and duration of benefit to first-line osimertinib therapy.
2. To decipher mechanisms of altered cell cycle progression and osimertinib-induced DNA damage that drive tumor progression.

We have assembled a collaborative, multi-disciplinary team of laboratory scientists and clinical investigators combined with cutting edge technologies to assess genetic and proteomic heterogeneity at the single cell level, access to a robust pipeline of patient samples, and strong preliminary data to support our studies. The proposed aims are expected to define the molecular determinants of tumor heterogeneity and to develop rational therapeutic strategies for delaying / overcoming resistance. These synergistic aims are highly relevant to the current landscape of EGFR-mutant lung cancer.

The current standard of care for ALK rearranged non-small cell lung cancer (NSCLC) is targeted therapy with ALK tyrosine kinase inhibitors (TKIs) which bind within the kinase domain of ALK. Due to the development of ALK resistance mutations or bypass activation of parallel or downstream signal pathways, patients inevitably relapse. Even with the introduction of second or third generation TKIs, median overall survival has been estimated to be 6.8 years . Recent advances in immunotherapy have demonstrated progress in lung cancer treatment. However, a majority of ALK positive NSCLC patients have yet to realize the benefits from checkpoint inhibitors or CAR-T cell therapies due to primary resistance or toxicity. Therefore, the development of new treatments for ALK-rearranged NSCLC is an area of critical need. We believe that an effective assault on ALK-driven tumors requires the implementation of creative strategies that allow us to expand our arsenal of drugs to target ALK beyond its kinase domain. To meet this challenge, we have developed and employed cysteine druggability mapping (CDM), a method to define the landscape of inhibitor-protein interactions in a cancer cell. CDM identifies cysteines that can bind to covalent drug-like molecules which serve as early prototypes on which to base cancer drug development programs.
Cysteine-reactive covalent inhibitors have already emerged as an effective approach to targeting protein kinases. Osimertinib, a third-generation epidermal growth factor receptor (EGFR) TKI covalent inhibitor, overcomes EGFR T790M resistance mutations following treatment with EGFR TKI, demonstrating significant improvements over standard EGFR TKIs when used as first line therapy. Importantly, targeting cysteines with covalent inhibitors massively expands the number of proteins which are targetable with small molecule inhibitors, including many targets previously considered undruggable.

We will use ALK fusion positive cancer cell lines, together with tumor tissues from patients who developed resistance mechanisms, to generate druggable cysteine profiles by CDM. Druggable proteins specific to ALK fusion positive samples, including EML4-ALK itself, will be prioritized in our screening for protein degraders from a library of over 2,500 advanced cysteine reactive compounds. After screening, we will assess the specificity of the compound hits and validate their anti-tumor activity using ALK fusion positive cell lines and patient derived ALK TKI-resistant cell lines. We consider the use of the novel, cutting-edge cysteine targeting technology to screen for potent therapeutic compounds as a valuable proposition to improve management of patients with ALK positive lung cancer, with the goal of transforming this cancer into a chronic condition.

With the development of increasingly potent and selective ALK tyrosine kinase inhibitors (TKIs), resistance is commonly mediated by ALK-independent mechanisms, such as bypass pathway activation. As bypass signaling dependencies vary across ALK-rearranged (ALK+) lung cancers, a multipronged approach is needed to broadly suppress potential bypass signals. We have identified MET activation in 20% of ALK+ tumors that are resistant to next-generation ALK TKIs. In preclinical studies, dual ALK/MET blockade re-sensitized MET-amplified ALK+ tumors to lorlatinib. MEK and SHP2 have recently been identified as important effector proteins downstream of several bypass signaling pathways. In ALK+ models harboring diverse bypass signals, pairing an ALK TKI with either a MEK or SHP2 inhibitor demonstrated broad anti-proliferative activity. Based on these findings, we will conduct a clinical trial to evaluate lorlatinib in combination with a MET, MEK, or SHP2 inhibitor for patients with ALK+ lung cancers with ALK-independent resistance. Serial tumor biopsies will be analyzed to characterize molecular determinants of response to combination therapy and identify expression changes induced by treatment. This innovative multi-arm trial will also assess whether dynamics of genetic alterations in plasma can be used to guide intensification of treatment from monotherapy to early combination therapy for patients who have not yet developed resistance to ALK-directed therapy. Collectively, the studies in this grant strive to develop diverse strategies for overcoming ALK-independent resistance, a heterogeneous molecular entity that undermines efficacy of current ALK therapeutics.

Several tyrosine kinase inhibitors (TKIs) are available for the treatment of ALK+ NSCLC, but resistance to all ALK TKIs (including alectinib, brigatinib, lorlatinib) has been reported and occurs through a variety of mechanisms. Once patients develop resistance to available ALK inhibitors, they are typically treated with cytotoxic chemotherapy rather than immunotherapy since response rates to PD-1 pathway inhibitors in this population are very low. However, our work in mouse models of ALK+ NSCLC has demonstrated that a vaccine generated against the rearranged portion of ALK can both prevent the development of ALK+ tumors and also more effectively treat them once they arise. We have also recently found evidence for anti-ALK immune responses in patients, as a subset of them generate spontaneous ALK autoantibodies.

We aim to translate our preclinical findings to a first-in-human clinical trial. Aim 1 will test the safety and efficacy of a novel ALK vaccine + an amphiphilic CpG toll-like receptor 9 agonist adjuvant, administered in combination with an ALK TKI or PD-1 inhibitor in patients with TKI-resistant ALK-positive NSCLC. In Aim 2, using liquid chromatography data-independent acquisition mass spectrometry, we will directly identify antigenic ALK peptides presented by MHC molecules on the surface of cancer cells from tumor biopsies obtained on trial participants. In Aim 3, we will study pre- and post-vaccination anti-ALK T-cell responses among trial participants by analyzing peripheral leukocytes and intratumoral immune cells. The overarching goal is to develop a vaccine that stimulates patients’ immune cells to recognize and eliminate ALK+ tumors.

Small cell lung cancer (SCLC) is an aggressive, neuroendocrine malignancy that accounts for approximately 15% of lung cancers. Despite recent advances, including the approval of immunotherapy combined with chemotherapy for frontline treatment, the median overall survival for patients with extensive-stage SCLC remains approximately one year. These dismal outcomes are attributable, in part, to the lack of approved biomarkers for SCLC, which, in turn, is due to the dearth of available tissue for molecular analysis. Using transcriptomic data, we have developed a gene signature that defines four distinct molecular subtypes of SCLC – each with unique biology and therapeutic vulnerabilities. These subtypes are driven by the presence (or absence) of three transcription factors (ASCL1 – SCLC-A; NEUROD1 – SCLC-N; POU2F3 – SCLC-P; and a fourth triple-negative subtype “Inflamed” – SCLC-I) that are largely mutually exclusive. Our preliminary data suggests that inter- and intra-tumoral heterogeneity among these subtypes drives response and resistance to standard and novel therapies for SCLC, including classes such as PARP inhibitors (PARPi) and immunotherapy. We propose applying bulk and single-cell molecular approaches to assign and longitudinally monitor subtype among our established cohort of patient-derived xenograft models treated with PARPi, as well as patient tumor biopsies from an investigator-initiated trial assessing a PARPi/immunotherapy combination. Based on our preliminary data, we hypothesize that SCLC-P and SCLC-I will preferentially benefit from PARP inhibitors and immunotherapy, respectively, and that subtype shifts detectable at the single-cell level along with accompanying shifts in the composition of the immune microenvironment, will govern response and resistance.

KRAS mutations are the most common activating mutation in lung cancer and are identified in 25-30% of cases of lung adenocarcinoma. The clinical activity of molecularly targeted therapy in KRAS mutant lung cancer has been limited, as most prior approaches focused on downstream targets of the MAPK pathway, an approach which was largely ineffective. KRAS G12C is the most common subtype of activating KRAS mutation in NSCLC. Identification of a binding pocket of KRAS G12C enabled the development of covalent inhibitors (Ostrem et al. Nature, 2013), which selectively bind to the mutant cysteine and spare wild type or other mutant KRAS proteins (hereafter referred to as G12Ci). These drugs bind only to GDP-bound (i.e. inactive) KRAS G12C and trap the oncoprotein in its inactive state by preventing nucleotide exchange (Lito et al. Science, 2016). Of this new class of agents, two drugs (MRTX849 and AMG510) have demonstrated efficacy in early phase clinical trials, representing a remarkable breakthrough in the management of KRAS G12C mutant lung cancer. Acknowledging the limited number of patients treated thus far, response rates of 40-50% have been reported, demonstrating clinical activity far superior to prior molecularly targeted approaches for the management of KRAS mutant lung cancer. Despite this major advance it is clear that resistance to therapy, both primary and acquired, will remain a challenge and may in part be due to upstream receptor tyrosine kinase (RTK) signaling. Understanding the mechanisms of resistance to this new class of therapy is of paramount importance to improve the efficacy and durability of response, and for identifying rational combinatorial therapies for future clinical trials.

Several ALK tyrosine kinase inhibitors (TKIs) are available for the treatment of ALK-rearranged non-small cell lung cancer (NSCLC), but resistance to all ALK TKIs (including crizotinib, ceritinib, alectinib, brigatinib, and lorlatinib) has been reported and occurs through a variety of mechanisms. Once patients develop resistance to available ALK inhibitors, they are typically treated with cytotoxic chemotherapy rather than immunotherapy since response rates to PD-1 pathway inhibitors in this population are very low. However, our work in mouse models of ALK-positive NSCLC has demonstrated that a vaccine generated against the rearranged portion of ALK can both prevent the development of ALK-positive tumors and also more effectively treat them once they arise. In addition, we have recently found evidence of an anti-ALK immune response in patients, as a subset of them generate spontaneous ALK autoantibodies.

To more deeply characterize anti-ALK immunologic responses in patients, we propose (Aim 1) to determine if the presence of anti-ALK antibodies correlates with progression-free and overall survival among patients with ALK-positive NSCLC, and (Aim 2) to directly identify and validate antigenic ALK peptides in different HLA-haplotype backgrounds. Antigenic peptides that are identified through these studies will serve as the basis for a therapeutic ALK peptide vaccine clinical trial, for which funding has been secured independently.

Background: Osimertinib and other “third-generation” EGFR TKIs effectively overcome T790M-mediated resistance in EGFR-mutant lung cancer, but our understanding of acquired resistance in this setting remains limited. Importantly, the intrinsic heterogeneity of resistant cancers leads to clinical challenges. We hypothesize that in-depth study of 3rd-generation TKI resistance and the impact of genetic heterogeneity on treatment response will lead to improved therapeutic options and clinical outcomes. Study Design: We will analyze tumor biopsies, ctDNA and autopsy specimens using a variety of techniques such as next-generation sequencing, digital droplet PCR and whole exome sequencing to characterize heterogeneity in resistant cancers. In addition, we will run a prospective clinical trial to test a novel therapeutic combination designed to minimize heterogeneity and delay progression. On-treatment biopsies and ctDNA collection will be used to study how resistance develops. Impact: This project will leverage our broad array of clinical specimens, well-established translational research program and preexisting industry and academic collaborations to characterize the heterogeneity of resistance in EGFR-mutant lung cancer. We hope to illuminate clinical and biologic implications of heterogeneity, with the ultimate goal of identifying more effective therapies for patients.

Immune-targeted agents have generated antitumor responses that are transforming cancer therapeutics in various tumors types including lung cancer. Despite the impressive responses observed, the majority of patients eventually experience therapeutic resistance after an initial response. Considering the increased cost to patients and health systems both financially and in terms of time spent in management of immune-related adverse effects it has become clear that success of immuno-oncology approaches depends on choosing patient populations most likely to benefit. Our proposed research addresses the urgent unmet clinical need to understand the underlying biology of response and develop molecular assays of response and resistance to immune targeted agents. We propose comprehensive genome-wide analyses of tumors and cell-free circulating DNA to examine how cancer genomes change during immune checkpoint blockade. Our approach is focused on discovery of mechanisms of resistance using a multidimensional strategy that combines comprehensive genomic analyses with functional assays of autologous T-cell reactivity. The underlying premise of this proposal is that through our comprehensive molecular analyses, we will identify changes in the genomic landscape of tumors during immune checkpoint blockade that mediate emergence of primary and acquired resistance to these therapies. Unravelling the genomic mechanisms through which cancer adapts to evade anti-tumor immune responses will be critical for future development of tailored cancer immunotherapy strategies. We envision that the results of the proposed research will be used to develop patient-specific immunotherapy approaches in lung cancer and will launch investigator-initiated clinical trials at Hopkins and other institutions.