Lineage tracing reveals clonal progenitors and long-term persistence of tumor-specific T cells during immune checkpoint blockade

Paired single-cell RNA and T cell receptor sequencing (scRNA/TCR-seq) has allowed for enhanced resolution of clonal T cell dynamics in cancer. Here, we report a scRNA/TCR-seq analysis of 187,650 T cells from 31 tissue regions, including tumor, adjacent normal tissues, and lymph nodes (LN), from three patients with non-small cell lung cancer after immune checkpoint blockade (ICB). Regions with viable cancer cells are enriched for exhausted CD8+ T cells, regulatory CD4+ T cells (Treg), and follicular helper CD4+ T cells (TFH). Tracking T cell clonotypes across tissues, combined with neoantigen specificity assays, reveals that TFH and tumor-specific exhausted CD8+ T cells are clonally linked to TCF7+SELL+ progenitors in tumor draining LNs, and progressive exhaustion trajectories of CD8+ T, Treg, and TFH cells with proximity to the tumor microenvironment. Finally, longitudinal tracking of tumor-specific CD8+ and CD4+ T cell clones reveals persistence in the peripheral blood for years after ICB therapy.

The ectonucleotidase CD39 identifies tumor-reactive CD8+ T cells predictive of immune checkpoint blockade efficacy in human lung cancer

Improved identification of anti-tumor T cells is needed to advance cancer immunotherapies. CD39 expression is a promising surrogate of tumor-reactive CD8+ T cells. Here, we comprehensively profiled CD39 expression in human lung cancer. CD39 expression enriched for CD8+ T cells with features of exhaustion, tumor reactivity, and clonal expansion. Flow cytometry of 440 lung cancer biospecimens revealed weak association between CD39+ CD8+ T cells and tumoral features, such as programmed death-ligand 1 (PD-L1), tumor mutation burden, and driver mutations. Immune checkpoint blockade (ICB), but not cytotoxic chemotherapy, increased intratumoral CD39+ CD8+ T cells. Higher baseline frequency of CD39+ CD8+ T cells conferred improved clinical outcomes from ICB therapy. Furthermore, a gene signature of CD39+ CD8+ T cells predicted benefit from ICB, but not chemotherapy, in a phase III clinical trial of non-small cell lung cancer. These findings highlight CD39 as a proxy of tumor-reactive CD8+ T cells in human lung cancer.

Abstract 704: Development of a patient-derived xenograft (PDX) modeling program to enable pediatric precision medicine

Background: Recapitulation of the full spectrum of genomic changes driving patient tumors have resulted in increased use of patient-derived xenograft (PDX) models in studies of basic cancer biology and preclinical drug development. Given the translational potential of PDXs and limited availability of pediatric cancer models, we established a PDX program to expand the existing collection of pediatric PDXs in the community and enable pre- and post-clinical studies.

Methods: PDX generation requests were integrated into clinical workflows to maximize identification of eligible patients for informed consent and tissue collection at Memorial Sloan Kettering Cancer Center. Methodologies for tissue procurement and cryopreservation were optimized to facilitate implantation into host immunodeficient mice and enable multi-institutional tissue exchange for model building. A bioinformatics pipeline was established to allow molecular validation of engrafted PDXs using a next-generation targeted gene panel (MSK-IMPACT) evaluating concordance based on acquired mutations, copy number alterations and clonal structure.

Results: Between November 2016 - October 2021, 379 PDX models were developed (265 distinct models) representing 69 discrete diagnoses. Sarcoma represents the most common model type (50 discrete osteosarcoma, 20 desmoplastic small round cell tumor, 14 Ewing sarcoma, 24 rhabdomyosarcoma, 2 CIC/DUX4 and 2 BCOR-rearranged sarcoma) followed by neuroblastoma (n=35), leukemia (n=44), and Wilms tumor (n=15). While the majority of PDXs were established from recurrent or metastatic tissue, 7 paired diagnostic/pre-therapy and post-therapy or relapse models were generated. Genomic characterization of PDXs demonstrate excellent concordance and recapitulation of single nucleotide variants (90%), structural (88%) and copy number variants (94%) between patient tumor and matched PDX. Discrepancies between matched patient/PDX pairs are due to sub-clonal heterogeneity in source tumors with clonal outgrowth in the PDX. Analysis of serial PDX passages also demonstrate stable recapitulation of the genomic profile. Establishment of a diverse PDX collection allowed preclinical evaluation of 10 targeted agents across a spectrum of pediatric tumors and provided the preclinical rationale for 3 investigator-initiated pediatric clinical trials.

Conclusions: Investment in the development of a phenotypically diverse and biologically faithful collection of pediatric PDX models enables the goals of precision medicine. Optimization of PDX workflows and methods has also enabled the development of a pediatric PDX consortium (PROXC - Pediatric Research in Oncology Xenografting Consortium) to further support the development of pre- and post-clinical studies for pediatric cancer.

Citation Format: Filemon S. Dela Cruz, Joseph G. McCarter, Daoqi You, Nancy Bouvier, Xinyi Wang, Kristina C. Guillan, Armaan H. Siddiquee, Katie B. Souto, Hongyan Li, Teng Gao, Dominik Glodzik, Daniel Diolaiti, Neerav N. Shukla, Joachim Silber, Umeshkumar K. Bhanot, Faruk Erdem Kombak, Diego F. Coutinho, Shanita Li, Juan E. Arango Ossa, Juan S. Medina-Martinez, Michael V. Ortiz, Emily K. Slotkin, Michael D. Kinnaman, Sameer F. Sait, Tara J. O'Donohue, Marissa Mattar, Maximiliano Meneses, Michael P. LaQuaglia, Todd E. Heaton, Justin T. Gerstle, Nicola Fabbri, Chelsey M. Burke, Irene M. Rodriquez-Sanchez, Christine A. Iacobuzio-Donahue, Julia L. Glade Bender, Ryan D. Roberts, Jason T. Yustein, Nino C. Rainusso, Brian D. Crompton, Elizabeth Stewart, Alejandro Sweet-Cordero, Leanne C. Sayles, Andrika D. Thomas, Michael H. Roehrl, Elisa de Stanchina, Elli Papaemmanuil, Andrew L. Kung. Development of a patient-derived xenograft (PDX) modeling program to enable pediatric precision medicine [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 704.

Defining and Targeting Esophagogastric Cancer Genomic Subsets With Patient-Derived Xenografts

PURPOSE

Comprehensive genomic profiling has defined key oncogenic drivers and distinct molecular subtypes in esophagogastric cancer; however, the number of clinically actionable alterations remains limited. To establish preclinical models for testing genomically driven therapeutic strategies, we generated and characterized a large collection of esophagogastric cancer patient-derived xenografts (PDXs).

MATERIALS AND METHODS

We established a biobank of 98 esophagogastric cancer PDX models derived from primary tumors and metastases. Clinicopathologic features of each PDX and the corresponding patient sample were annotated, including stage at diagnosis, treatment history, histology, and biomarker profile. To identify oncogenic DNA alterations, we analyzed and compared targeted sequencing performed on PDX and parent tumor pairs. We conducted xenotrials in genomically defined models with oncogenic drivers.

RESULTS

From April 2010 to June 2019, we implanted 276 patient tumors, of which 98 successfully engrafted (35.5%). This collection is enriched for PDXs derived from patients with human epidermal growth factor receptor 2-positive esophagogastric adenocarcinoma (62 models, 63%), the majority of which were refractory to standard therapies including trastuzumab. Factors positively correlating with engraftment included advanced stage, metastatic origin, intestinal-type histology, and human epidermal growth factor receptor 2-positivity. Mutations in TP53 and alterations in receptor tyrosine kinases (ERBB2 and EGFR), RAS/PI3K pathway genes, cell-cycle mediators (CDKN2A and CCNE1), and CDH1 were the predominant oncogenic drivers, recapitulating clinical tumor sequencing. We observed antitumor activity with rational combination strategies in models established from treatment-refractory disease.

CONCLUSION

The Memorial Sloan Kettering Cancer Center PDX collection recapitulates the heterogeneity of esophagogastric cancer and is a powerful resource to investigate mechanisms driving tumor progression, identify predictive biomarkers, and develop therapeutic strategies for molecularly defined subsets of esophagogastric cancer.

Pharmacologically controlling protein-protein interactions through epichaperomes for therapeutic vulnerability in cancer

Cancer cell plasticity due to the dynamic architecture of interactome networks provides a vexing outlet for therapy evasion. Here, through chemical biology approaches for systems level exploration of protein connectivity changes applied to pancreatic cancer cell lines, patient biospecimens, and cell- and patient-derived xenografts in mice, we demonstrate interactomes can be re-engineered for vulnerability. By manipulating epichaperomes pharmacologically, we control and anticipate how thousands of proteins interact in real-time within tumours. Further, we can essentially force tumours into interactome hyperconnectivity and maximal protein-protein interaction capacity, a state whereby no rebound pathways can be deployed and where alternative signalling is supressed. This approach therefore primes interactomes to enhance vulnerability and improve treatment efficacy, enabling therapeutics with traditionally poor performance to become highly efficacious. These findings provide proof-of-principle for a paradigm to overcome drug resistance through pharmacologic manipulation of proteome-wide protein-protein interaction networks.

Signatures of plasticity, metastasis, and immunosuppression in an atlas of human small cell lung cancer

Small cell lung cancer (SCLC) is an aggressive malignancy that includes subtypes defined by differential expression of ASCL1, NEUROD1, and POU2F3 (SCLC-A, -N, and -P, respectively). To define the heterogeneity of tumors and their associated microenvironments across subtypes, we sequenced 155,098 transcriptomes from 21 human biospecimens, including 54,523 SCLC transcriptomes. We observe greater tumor diversity in SCLC than lung adenocarcinoma, driven by canonical, intermediate, and admixed subtypes. We discover a PLCG2-high SCLC phenotype with stem-like, pro-metastatic features that recurs across subtypes and predicts worse overall survival. SCLC exhibits greater immune sequestration and less immune infiltration than lung adenocarcinoma, and SCLC-N shows less immune infiltrate and greater T cell dysfunction than SCLC-A. We identify a profibrotic, immunosuppressive monocyte/macrophage population in SCLC tumors that is particularly associated with the recurrent, PLCG2-high subpopulation.

Tim-4+ cavity-resident macrophages impair anti-tumor CD8+ T cell immunity

Immune checkpoint blockade (ICB) has been a remarkable clinical advance for cancer; however, the majority of patients do not respond to ICB therapy. We show that metastatic disease in the pleural and peritoneal cavities is associated with poor clinical outcomes after ICB therapy. Cavity-resident macrophages express high levels of Tim-4, a receptor for phosphatidylserine (PS), and this is associated with reduced numbers of CD8+ T cells with tumor-reactive features in pleural effusions and peritoneal ascites from patients with cancer. We mechanistically demonstrate that viable and cytotoxic anti-tumor CD8+ T cells upregulate PS and this renders them susceptible to sequestration away from tumor targets and proliferation suppression by Tim-4+ macrophages. Tim-4 blockade abrogates this sequestration and proliferation suppression and enhances anti-tumor efficacy in models of anti-PD-1 therapy and adoptive T cell therapy in mice. Thus, Tim-4+ cavity-resident macrophages limit the efficacy of immunotherapies in these microenvironments.

Abstract 2811: Temporal modulation of antigen availability to improve antibody-tumor binding and therapeutic efficacy

Statins are low-cost cholesterol-depleting drugs used to treat patients with hypercholesterolemia. Preclinically, statins modulate caveolae-mediated endocytosis. Membrane receptors defined as tumor biomarkers and therapeutic targets are often internalized by an endocytic pathway. Indeed, receptor endocytosis and recycling are dynamic mechanisms that can affect receptor density at the cell surface. In therapies using monoclonal antibodies, a downregulation in receptor density at the cell surface decreases antibody binding to the extracellular domain of the membrane receptor. Here, we used immunoPET to demonstrate that statins can temporally modulate human epidermal growth factor receptor 2 (HER2), epidermal growth factor receptor (EGFR), and prostate-specific membrane antigen (PSMA) receptor density at the tumor cell surface for binding therapeutic monoclonal antibodies and antibody-drug conjugates. In xenografts and patient samples, we found that tumors with high CAV1 expression localized less receptor at the cell membrane and these features decreased antibody uptake and efficacy. In cultured cells and tumor xenografts, statins temporally depleted CAV1 expression in ways than enhanced tumors' avidity for anti-HER2, anti-EGFR, and anti-PSMA antibodies. In HER2- and EGFR-expressing xenografts, treating mice with a statin accelerated and increased 89Zr-labeled antibody accumulation at the tumor site. The hydrophilic rosuvastatin demonstrated a lower ability to enhance antibody binding to tumors when compared with the lipophilic lovastatin and simvastatin. 89Zr-labeled huJ591 tumor accumulation was higher in PSMA-expressing tumors of statin-treated mice when compared with tumors of saline-treated mice at later time-points of antibody accumulation (24 h and 48 h), but the values were similar at 4 h and 8 h. These results suggest that treating mice with statins increases, but does not accelerate, anti-PSMA antibody accumulation. Anti-DLL3 antibody accumulation was similar in saline versus statin treated tumors. Retrospective data demonstrated that patients with HER2+/CAV1HIGH(29.4% of total HER2+gastric tumors) have a significantly worse overall survival than those expressing low CAV1. Kaplan-Meier analyses of statin use and HER2+gastric cancer disease outcome in patients treated with trastuzumab suggested that patients without statin treatment have worse survival than patients treated with a statin. Additionally, statins synergize with therapeutic antibodies to decrease oncogenic signaling pathways.Our data suggest that acute statin treatment with appropriate pharmacokinetics/pharmacodynamics are potential adjuvants for specific antibody-targeted therapies.

Citation Format: Patricia Pereira, Komal Mandleywala, Ashwin Ragupathi, Marissa Mattar, Sébastien Monette, Yelena Janjigian, Jason Lewis. Temporal modulation of antigen availability to improve antibody-tumor binding and therapeutic efficacy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 2811.

Abstract 956: The HER2×HER3 bi-specific antibody Zenocutuzumab is effective at blocking growth of tumors driven by NRG1 gene fusions

Fusions involving the neuregulin 1 gene (NRG1) occur at low frequency in pancreatic, lung, and other cancers. NRG1 fusion oncoproteins bind to HER3, leading to heterodimerization with HER2 and potent activation of downstream signaling mainly via the PI3K-AKT pathway. Zenocutuzumab (Zeno, MCLA-128), an ADCC-enhanced anti-HER2×HER3 bi-specific antibody, uniquely ‘docks' on HER2, to position the antibody and subsequently ‘block' NRG1 from interacting with HER3, effectively preventing HER2:HER3 heterodimerization and downstream signaling. Our goal in this study was to evaluate the efficacy of Zeno in preclinical models of NRG1 fusion-positive cancers.

We tested Zeno in a panel of isogenic and patient-derived cell line and xenograft (PDX) models of lung, breast and pancreatic cancers. Cell lines either expressed an NRG1 fusion endogenously (MDA-MB-175-VII, DOC4-NRG1) or by lentiviral transfer of cDNAs (ATP1B1-NRG1 and SLC3A2-NRG1 in H6c7 pancreatic ductal cell line; CD74-NRG1 and VAMP2-NRG1 in immortalized human bronchial epithelial cells; and DOC4-NRG1 in MCF7 breast cancer cells). PDX models were generated from NSCLC samples harboring CD74-NRG1 (ST3204) or SLC3A2-NRG1 (LUAD-0061AS3) fusions and from a high grade serous ovarian cancer harboring a CLU-NRG1 fusion (OV-10-0050). Zeno treatment of NRG1 fusion-expressing breast, pancreatic, and lung cancer cell lines resulted in dose-dependent reduction of growth and abrogated phosphorylation of HER3, HER4, AKT, p70S6 kinase and STAT3 in all cell lines tested. Phosphorylation of HER2, EGFR and MEK/ERK was inhibited, albeit with some variation, in a cell line-specific manner. Growth of isogenic control cell lines without NRG1 fusion was not significantly altered. In breast and lung cancer cell lines, Zeno treatment down-regulated cyclin D1 expression and induced expression of the negative cell cycle regulators P21 or P27. Evidence of apoptosis activation (cleaved PARP, expression of BIM and PUMA) was also observed in cells exposed to Zeno. Treatment of mice bearing LUAD-0061AS3, ST3204 and OV-10-0050 PDX tumors (2.5, 8, 25 mg/kg, QW) caused a dose-dependent inhibition of tumor growth, with tumor shrinkage observed at higher doses. Finally, we assessed the ability of Zeno to induce antibody-dependent cellular cytotoxicity using a chromium release assay and peripheral blood mononuclear cells. Zeno induced significant cytotoxicity in MDA-MB-175-VII cells while a non-ADCC enhanced, non-specific IgG had no effect.

Here we show that Zeno effectively blocks the growth of NRG1 fusion-positive cell line and xenograft models of tumors arising from lung, pancreas and other organs, and these results support the continued development of Zeno to treat patients with this molecularly defined subset of cancers.

Citation Format: Igor Odintsov, Inna Khodos, Madelyn Espinosa-Cotton, Allan J. Lui, Marissa Mattar, Alison M. Schram, Ron C. Schackmann, Jeroen Lammerts van Bueren, Cecile A. Geuijen, Elisa de Stanchina, Marc Ladanyi, Romel Somwar. The HER2×HER3 bi-specific antibody Zenocutuzumab is effective at blocking growth of tumors driven by NRG1 gene fusions [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 956.

Signatures of plasticity and immunosuppression in a single-cell atlas of human small cell lung cancer

8509

Background: Small cell lung cancer (SCLC) is an aggressive malignancy that includes subtypes defined by differential expression of ASCL1, NEUROD1, and POU2F3 (SCLC-A, -N, and -P, respectively), which are associated with distinct therapeutic vulnerabilities. The emerging consensus on SCLC subtypes has led to new questions, such as whether subtypes are associated with different disease stages, metastatic potential, or immune microenvironments; whether there is plasticity between subtypes; and whether novel SCLC phenotypes exist. Single cell RNA sequencing (scRNA-seq) offers a unique opportunity to address these questions by dissecting intratumoral transcriptional heterogeneity and the surrounding tumor microenvironment (TME). However, efforts to apply this technology to human SCLC tumors have been limited, as these tumors are infrequently resected. Methods: We have optimized protocols to process both surgical resections and biopsies to construct the first single-cell atlas of SCLC patient tumors (N = 21), with comparative lung adenocarcinoma (LUAD) and normal lung data. We leverage computational methods including diffusion maps and non-negative matrix factorization to perform a deep annotation of SCLC phenotypes and the surrounding immune TME. We perform validation experiments using flow cytometry, Vectra, and immunohistochemistry in independent SCLC cohorts, as well as genetic manipulation in preclinical SCLC models. Results: Our data reveals substantial transcriptional heterogeneity in SCLC both within and across tumors and confirms a pro-metastatic gene program in SCLC-N subtype characterized by epithelial-mesenchymal transformation and axonogenesis. Beyond known subtypes, we discover a PLCG2-high tumor cell population with stem-like, pro-metastatic features that recurs across subtypes and predicts significantly worse overall survival. Manipulation of PLCG2 expression in cells confirms correlation with key metastatic markers. Treatment and subtype are associated with substantial phenotypic changes in the SCLC immune microenvironment, with greater T-cell dysfunction in SCLC-N than SCLC-A. Moreover, the recurrent, PLCG2-high subclone is associated with exhausted CD8+ T-cells and a pro-fibrotic, immunosuppressive monocyte/macrophage population, suggesting possible tumor-immune coordination to promote metastasis. Conclusions: This atlas of SCLC illustrates how canonical subtypes and a novel PLCG2-high recurrent tumor subclone enlist diverse gene programs to create tumor heterogeneity and facilitate metastasis in a profoundly immunosuppressed TME. Our dataset provides further insight into tumor and immune biology in SCLC at single-cell resolution, with potential implications for design of novel targeted therapies and immunotherapeutic approaches.

Impact of tumor heterogeneity and microenvironment in identifying neoantigens in a patient with ovarian cancer

Identification of neoepitopes as tumor-specific targets remains challenging, especially for cancers with low mutational burden, such as ovarian cancer. To identify mutated human leukocyte antigen (HLA) ligands as potential targets for immunotherapy in ovarian cancer, we combined mass spectrometry analysis of the major histocompatibility complex (MHC) class I peptidomes of ovarian cancer cells with parallel sequencing of whole exome and RNA in a patient with high-grade serous ovarian cancer. Four of six predicted mutated epitopes capable of binding to HLA-A*02:01 induced peptide-specific T cell responses in blood from healthy donors. In contrast, all six peptides failed to induce autologous peptide-specific response by T cells in peripheral blood or tumor-infiltrating lymphocytes from ascites of the patient. Surprisingly, T cell responses against a low-affinity p53-mutant Y220C epitope were consistently detected in the patient with either unprimed or in vitro peptide-stimulated T cells even though the patient's primary tumor did not bear this mutation. Our results demonstrated that tumor heterogeneity and distinct immune microenvironments within a patient should be taken into consideration for identification of immunogenic neoantigens. T cell responses to a driver gene-derived p53 Y220C mutation in ovarian cancer warrant further study.

Genomic characterization of a comprehensive collection of esophagogastric cancer patient-derived xenografts

233

Background: Large-scale sequencing has identified multiple oncogenic drivers and molecularly defined subsets in esophagogastric cancer. However, besides therapies for HER2-positive and microsatellite instability-high disease, few genomic biomarker-driven treatments are currently approved. Patient-derived xenografts (PDXs) have emerged as promising preclinical models that capture the heterogeneity and biology of human tumors. Therefore, we established a comprehensive collection of esophagogastric cancer PDXs and performed next-generation sequencing (NGS) to genomically characterize these models. Methods: Starting in 2010, we developed an ongoing program for generating esophagogastric cancer PDXs from fresh tumor specimens that are acquired from surgical resections or biopsies and implanted into NOD scid gamma (NSG) mice either subcutaneously into flanks or orthotopically into the gastric wall. We reviewed clinical and pathologic characteristics of patients from whom established PDXs were derived, including stage, histology, HER2 status and treatment history. To identify oncogenic DNA alterations, NGS was performed on PDX material using MSK-IMPACT, a capture-based NGS platform. Results: From April 2010 to August 2018, we implanted 270 tumor samples, of which 112 were successfully engrafted (41.4%) including 57 gastric adenocarcinomas, 25 gastroesophageal junction adenocarcinomas, 23 esophageal adenocarcinomas, 4 squamous cell carcinomas and 3 small cell/high-grade neuroendocrine tumors. PDXs were generated from both primary tumors (n = 67, 59.8%) and metastases (n = 45, 40.2%), with many PDXs established from patients with metastatic disease who had progressed on standard therapy (n = 50, 44.6%). In addition, a large number of PDXs were derived from patients initially diagnosed with HER2-positive esophagogastric adenocarcinoma (n = 68, 60.7%). NGS of these PDXs demonstrated frequent alterations in TP53, receptor tyrosine kinases ( ERBB2, EGFR), cell-cycle mediators ( CDK12, CCNE1, CCND3) and histone methyltransferases ( KMT2C, KMT2D), consistent with clinical sequencing. Conclusions: Comprehensive molecular profiling demonstrates that esophagogastric cancer PDXs recapitulate the genomic heterogeneity and tumor biology of patients. This panel represents one of the largest described collections of esophagogastric cancer PDXs and serves as a powerful platform to investigate mechanisms driving tumor growth and metastasis, identify predictive biomarkers for treatment responsiveness and develop novel genomically-driven therapeutic strategies.

HER2-Mediated Internalization of Cytotoxic Agents in ERBB2 Amplified or Mutant Lung Cancers

Amplification of and oncogenic mutations in ERBB2, the gene encoding the HER2 receptor tyrosine kinase, promote receptor hyperactivation and tumor growth. Here we demonstrate that HER2 ubiquitination and internalization, rather than its overexpression, are key mechanisms underlying endocytosis and consequent efficacy of the anti-HER2 antibody-drug conjugates (ADC) ado-trastuzumab emtansine (T-DM1) and trastuzumab deruxtecan (T-DXd) in lung cancer cell lines and patient-derived xenograft models. These data translated into a 51% response rate in a clinical trial of T-DM1 in 49 patients with ERBB2-amplified or -mutant lung cancers. We show that cotreatment with irreversible pan-HER inhibitors enhances receptor ubiquitination and consequent ADC internalization and efficacy. We also demonstrate that ADC switching to T-DXd, which harbors a different cytotoxic payload, achieves durable responses in a patient with lung cancer and corresponding xenograft model developing resistance to T-DM1. Our findings may help guide future clinical trials and expand the field of ADC as cancer therapy. SIGNIFICANCE: T-DM1 is clinically effective in lung cancers with amplification of or mutations in ERBB2. This activity is enhanced by cotreatment with irreversible pan-HER inhibitors, or ADC switching to T-DXd. These results may help address unmet needs of patients with HER2-activated tumors and no approved targeted therapy.See related commentary by Rolfo and Russo, p. 643.This article is highlighted in the In This Issue feature, p. 627.

MAPK Pathway Alterations Correlate with Poor Survival and Drive Resistance to Therapy in Patients with Lung Cancers Driven by ROS1 Fusions

PURPOSE

ROS1 tyrosine kinase inhibitors (TKI) provide significant benefit in lung adenocarcinoma patients with ROS1 fusions. However, as observed with all targeted therapies, resistance arises. Detecting mechanisms of acquired resistance (AR) is crucial to finding novel therapies and improve patient outcomes.

EXPERIMENTAL DESIGN

ROS1 fusions were expressed in HBEC and NIH-3T3 cells either by cDNA overexpression (CD74/ROS1, SLC34A2/ROS1) or CRISPR-Cas9-mediated genomic engineering (EZR/ROS1). We reviewed targeted large-panel sequencing data (using the MSK-IMPACT assay) patients treated with ROS1 TKIs, and genetic alterations hypothesized to confer AR were modeled in these cell lines.

RESULTS

Eight of the 75 patients with a ROS1 fusion had a concurrent MAPK pathway alteration and this correlated with shorter overall survival. In addition, the induction of ROS1 fusions stimulated activation of MEK/ERK signaling with minimal effects on AKT signaling, suggesting the importance of the MAPK pathway in driving ROS1 fusion-positive cancers. Of 8 patients, 2 patients harbored novel in-frame deletions in MEK1 (MEK1delE41_L54) and MEKK1 (MEKK1delH907_C916) that were acquired after ROS1 TKIs, and 2 patients harbored NF1 loss-of-function mutations. Expression of MEK1del or MEKK1del, and knockdown of NF1 in ROS1 fusion-positive cells activated MEK/ERK signaling and conferred resistance to ROS1 TKIs. Combined targeting of ROS1 and MEK inhibited growth of cells expressing both ROS1 fusion and MEK1del.

CONCLUSIONS

We demonstrate that downstream activation of the MAPK pathway can mediate of innate acquired resistance to ROS1 TKIs and that patients harboring ROS1 fusion and concurrent downstream MAPK pathway alterations have worse survival. Our findings suggest a treatment strategy to target both aberrations.

Regenerative lineages and immune-mediated pruning in lung cancer metastasis

Developmental processes underlying normal tissue regeneration have been implicated in cancer, but the degree of their enactment during tumor progression and under the selective pressures of immune surveillance, remain unknown. Here, we show that human primary lung adenocarcinomas are characterized by the emergence of regenerative cell types typically seen in response to lung injury, and by striking infidelity amongst transcription factors specifying most alveolar and bronchial epithelial lineages. In contrast, metastases are enriched for key endoderm and lung-specifying transcription factors, SOX2 and SOX9, and recapitulate more primitive transcriptional programs spanning stem-like to regenerative pulmonary epithelial progenitor states. This developmental continuum mirrors the progressive stages of spontaneous outbreak from metastatic dormancy in a mouse model and exhibits SOX9-dependent resistance to Natural Killer (NK) cells. Loss of developmental stage-specific constraint in macrometastases triggered by NK cell depletion suggests a dynamic interplay between developmental plasticity and immune-mediated pruning during metastasis.

LPTO-04. GENERATION AND CHARACTERIZATION OF PATIENT-DERIVED PRECLINICAL MODELS FROM TUMOR CELLS ISOLATED FROM CEREBROSPINAL FLUID

BACKGROUND: CNS metastases occur in 20–50% of lung cancer patients during their disease; leptomeningeal disease (LMD) representing 5–8%, classically carries a poor prognosis with a median overall survival ranging from 1–11 months. There is a paucity of patient-derived preclinical disease models using tumor cells isolated from the CSF. Models that faithfully recapitulate the biology of CNS tumors would offer new insights into the biology of the disease as well as provide the basis for developing more effective therapy. METHODS: To create more representative preclinical models to study LMD we isolated tumor cells from CSF of 5 patients with cytologically proven LMD and implanted the cells into the subcutaneous flank of immune-compromised mice. Where possible, cell lines were also generated from PDX tissues. Models were characterized by next generation sequencing (NGS), growth rates, expression of driver oncogenes and sensitivity to small molecule inhibitors. RESULTS: To date, one PDX (LUAD-0048A) and cell line model were successfully derived from CSF samples (NSCLC patient with MET amplification) and 4 are pending. MET amplification and mRNA over-expression were confirmed by quantitative PCR in the PDX tissue and the cell line. Western blot analysis indicated that over-expressed MET was phosphorylated in both PDX tissue and cell line. These results were confirmed by immunohistochemistry. Growth of LUAD-0048A cells were unaffected by 3 MET inhibitors (crizotinib, cabozantinib, glesatinib). Similarly, MET inhibitors did not induce apoptosis in the cells. CONCLUSION: LMD represents an aggressive metastatic event in lung cancer patients. Here we were able to successfully establish a PDX from the CSF of a patient with LMD and trial targeted therapies in vivo. Translational collaborations where patients with clinical suspicion of LMD undergo CSF sampling, NGS/ctDNA analysis, and PDX modeling are crucial in improving our understanding of this metastatic compartment and investigating novel treatment paradigms.

Abstract LB-118: Resistance to TRK inhibition mediated by convergent MAP kinase pathway activation

Background: TRK inhibition is now standard of care for advanced pediatric and adult patients (pts) with TRK fusion solid tumors, regardless of origin. To date, TRK kinase domain mutations are the only known resistance mechanism, and next-generation TRK inhibitors active against these mutations such as LOXO-195 are being developed. We reasoned some pts will develop TRK-independent resistance and hypothesized that these pts will require unique therapeutic approaches.

Methods: Paired tumor biopsies and serial cell-free DNA (cfDNA) prospectively collected from pts with TRK fusion-positive cancers treated with first- and next-generation TRK inhibitors before treatment and at progression were sequenced. In parallel, pt-derived and engineered models were analyzed.

Results: Alterations involving upstream non-TRK receptor kinases and downstream MAPK pathway members were initially identified in tumors from 3 TRK fusion-positive gastrointestinal (GI) cancer pts who developed resistance to TRK inhibitors. Pt 1 with CTRC-NTRK1 pancreatic cancer developed temporally distinct emergent BRAF V600E and KRAS G12D mutations. Pt 2 with LMNA-NTRK1 colorectal cancer developed temporally distinct KRAS G12A and G12D mutations. Pt 3 with PLEKHA6-NTRK1 cholangiocarcinoma developed focal MET amplification. Phenocopying these clinical observations, pt-derived xenografts and primary cell lines developed BRAF and KRAS mutations following chronic TRK inhibition. Consistently, ectopic expression of these alterations conferred resistance to TRK inhibitors. Given that all 3 index pts had GI cancers, we expanded serial cfDNA sequencing to 5 additional TRK fusion-positive GI disease, identifying 3 with emergent MAPK alterations at progression, bringing the overall frequency of acquired MAPK alterations in GI cancers analyzed to 75% (6/8). To further evaluate whether these emergent alterations induced functional dependence on ERK signaling, pts 1-3 were treated with agents targeting these emergent alterations (dabrafenib + trametinib, LOXO-195 + trametinib, and LOXO-195 + crizotinib, respectively). Pt 1 achieved transient tumor regression, followed by outgrowth of KRAS-mutant disease. Pt 3 achieved a 4.5 months tumor regression. Sequencing at progression in pt 3 identified multiple acquired MET point mutations known to interfere with crizotinib binding.

Conclusions: These data suggest that a subset of TRK fusion-positive cancers will develop off-target mechanisms of resistance to TRK inhibition. Relative to other TRK fusion-positive tumors, GI cancers may have a higher propensity for developing these bypass alterations that demonstrate remarkable convergence on ERK signaling. A portion of these mechanisms may be managed with simultaneous targeting of the TRK and MAPK pathways, although additional modeling is required to determine if upfront treatment would confer more durable responses.

Citation Format: Emiliano Cocco, Amanda Kulick, Sandra Misale, Rona Yaeger, Pedram Razavi, Helen H. Won, Ryan Ptashkin, Jaclyn F. Hechtman, Eneda Toska, James Cownie, Romel Somwar, Sophie Shifman, Marissa Mattar, S Duygu Selçuklu, Aliaksandra Samoila, Sean Guzman, Brian B. Tuch, Kevin Ebata, Elisa de Stanchina, Rebecca J. Nagy, Richard B. Lanman, Michael F. Berger, Marc Ladanyi, David M. Hyman, Alexander Drilon, Maurizio Scaltriti, Alison M. Schram. Resistance to TRK inhibition mediated by convergent MAP kinase pathway activation [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr LB-118.

EGFR and MET Amplifications Determine Response to HER2 Inhibition in ERBB2-Amplified Esophagogastric Cancer

The anti-HER2 antibody trastuzumab is standard care for advanced esophagogastric (EG) cancer with ERBB2 (HER2) amplification or overexpression, but intrinsic and acquired resistance are common. We conducted a phase II study of afatinib, an irreversible pan-HER kinase inhibitor, in trastuzumab-resistant EG cancer. We analyzed pretreatment tumor biopsies and, in select cases, performed comprehensive characterization of postmortem metastatic specimens following acquisition of drug resistance. Afatinib response was associated with coamplification of EGFR and ERBB2. Heterogeneous 89Zr-trastuzumab PET uptake was associated with genomic heterogeneity and mixed clinical response to afatinib. Resistance to afatinib was associated with selection for tumor cells lacking EGFR amplification or with acquisition of MET amplification, which could be detected in plasma cell-free DNA. The combination of afatinib and a MET inhibitor induced complete tumor regression in ERBB2 and MET coamplified patient-derived xenograft models established from a metastatic lesion progressing on afatinib. Collectively, differential intrapatient and interpatient expression of HER2, EGFR, and MET may determine clinical response to HER kinase inhibitors in ERBB2-amplified EG cancer. SIGNIFICANCE: Analysis of patients with ERBB2-amplified, trastuzumab-resistant EG cancer who were treated with the HER kinase inhibitor afatinib revealed that sensitivity and resistance to therapy were associated with EGFR/ERBB2 coamplification and MET amplification, respectively. HER2-directed PET imaging and cell-free DNA sequencing could help guide strategies to overcome the emergence of resistant clones.See related commentary by Klempner and Catenacci, p. 166.This article is highlighted in the In This Issue feature, p. 151.

Activation of KRAS Mediates Resistance to Targeted Therapy in MET Exon 14-mutant Non-small Cell Lung Cancer

PURPOSE

MET exon 14 splice site alterations that cause exon skipping at the mRNA level (METex14) are actionable oncogenic drivers amenable to therapy with MET tyrosine kinase inhibitors (TKI); however, secondary resistance eventually arises in most cases while other tumors display primary resistance. Beyond relatively uncommon on-target MET kinase domain mutations, mechanisms underlying primary and acquired resistance remain unclear.

EXPERIMENTAL DESIGN

We examined clinical and genomic data from 113 patients with lung cancer with METex14. MET TKI resistance due to KRAS mutation was functionally evaluated using in vivo and in vitro models.

RESULTS

Five of 113 patients (4.4%) with METex14 had concurrent KRAS G12 mutations, a rate of KRAS cooccurrence significantly higher than in other major driver-defined lung cancer subsets. In one patient, the KRAS mutation was acquired post-crizotinib, while the remaining 4 METex14 patients harbored the KRAS mutation prior to MET TKI therapy. Gene set enrichment analysis of transcriptomic data from lung cancers with METex14 revealed preferential activation of the KRAS pathway. Moreover, expression of oncogenic KRAS enhanced MET expression. Using isogenic and patient-derived models, we show that KRAS mutation results in constitutive activation of RAS/ERK signaling and resistance to MET inhibition. Dual inhibition of MET or EGFR/ERBB2 and MEK reduced growth of cell line and xenograft models.

CONCLUSIONS

KRAS mutation is a recurrent mechanism of primary and secondary resistance to MET TKIs in METex14 lung cancers. Dual inhibition of MET or EGFR/ERBB2 and MEK may represent a potential therapeutic approach in this molecular cohort.

Abstract 1826: Oncogenic KRAS mediates resistance to MET targeted therapy in non-small cell lung cancer (NSCLC) with MET mutations that induce exon14 skipping

Mutations in MET that induce skipping of exon 14 and lead to reduced ubiquitin ligase-mediated turnover of this receptor tyrosine kinase (RTK) are detected in 3-4% of non-small cell lung cancer (NSCLC), approaching the prevalence of ALK-rearranged lung cancers. Preclinical and clinical studies have revealed that MET exon14 alterations are actionable oncogenic drivers that are amenable to therapy with MET kinase inhibitors such as crizotinib. However, similar to most kinase-driven cancers, despite initial benefit, acquired resistance to therapy is inevitable. Next-generation sequencing (MSK-IMPACT 468 gene panel) was performed on samples from 81 NSCLC patients with MET exon14 alterations, including 7 with paired pre- and post-treatment tumor samples. A concurrent KRAS G12 mutation was identified in 5 patients. In 4 of these patients, the KRAS mutation was present prior to receiving crizotinib. The KRAS mutation was acquired post-crizotinib in the remaining patient. These findings implicate KRAS activation as a potential mechanism of acquired resistance. Using isogenic and patient-derived in vitro and in vivo models harboring MET exon14 skipping alteration, we confirmed that the KRAS mutation results in constitutive activation of RAS/ERK signaling and cells expressing both MET exon14 skipping and KRAS mutations are refractory to MET inhibition. Dual inhibition of MET and MEK with crizotinib and trametinib, respectively, has an additive effect in cell line and xenograft models. Whereas concurrent KRAS mutation is an extremely rare event in EGFR- and ALK-driven NSCLC, our findings confirm KRAS mutation as a recurrent mechanism of primary or secondary resistance to MET-directed therapies in lung cancers harboring MET exon14 alterations. We provide a new potential therapeutic strategy for NSCLC patients with both MET exon14 alterations and KRAS mutations.

Citation Format: Ken Suzawa, Michael D. Offin, Christopher Kurzatkowski, Daniel Liu, Morana Vojnic, Roger S. Smith, Marissa Mattar, Inna Khodos, Elisa de Stanchina, Joshua K. Sabari, William W. Lockwood, Alexander E. Drilon, Marc Ladanyi, Romel Somwar. Oncogenic KRAS mediates resistance to MET targeted therapy in non-small cell lung cancer (NSCLC) with MET mutations that induce exon14 skipping [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 1826.

Establishing and Maintaining an Extensive Library of Patient-Derived Xenograft Models

Patient-derived xenograft (PDX) models have recently emerged as a highly desirable platform in oncology and are expected to substantially broaden the way in vivo studies are designed and executed and to reshape drug discovery programs. However, acquisition of patient-derived samples, and propagation, annotation and distribution of PDXs are complex processes that require a high degree of coordination among clinic, surgery and laboratory personnel, and are fraught with challenges that are administrative, procedural and technical. Here, we examine in detail the major aspects of this complex process and relate our experience in establishing a PDX Core Laboratory within a large academic institution.

Role of ERBB signaling in RET-rearranged lung cancer and contribution of EGFR amplification to cabozantinib resistance

11583

Background: Lung cancers driven by oncogenic RET fusions have lower response rates to targeted monotherapy such as cabozantinib (28%) relative to response rates typically observed in ALK- or ROS1- rearranged lung adenocarcinomas (60-80%). Methods: To identify targetable co-dependencies or cooperating pathways for RET fusion-positive lung cancers, we performed high-throughput chemical and genetic screens to find FDA-approved drugs or genes that when inhibited, would synergize with cabozantinib in RET fusion-positive lung cancer cell lines. In addition we performed NGS of a pair of pre-treatment and post-cabozantinib progression samples. Results: We identified EGFR siRNAs and anti-EGFR drugs as synergistic with cabozantinib. Combinations of drugs that target EGFR (cetuximab, afatinib, erlotinib, gefitinib, neratinib) and RET (cabozantinib, CEP-32496, lenvatinib, vandetanib) were more effective at reducing growth of RET cell lines than any single agent in vitro and in xenograft models. Cabozantinib treatment of RET fusion-positive cell lines inhibited EGFR and RET phosphorylation, an observation not seen in RET wild-type cell lines. Co-immunoprecipitation studies reveal that RET and EGFR interact. Ectopic expression of CCDC6-RET in NIH-3T3 or human bronchial epithelial cells resulted in upregulation of multiple ERBB receptors and ligands (not seen in a ROS1 fusion-positive cell line) and a concomitant increase in EGFR stability. Treatment with ERBB pathway ligands or overexpression of EGFR decreased sensitivity to cabozantinib in two RET fusion-positive cell lines. Finally, sequencing of a pair of pre-treatment and post-progression samples from a lung cancer patient treated with cabozantinib revealed acquired amplification of EGFR in the latter sample. Conclusions: Taken together, these results suggest that the tumorigenic potential of RET fusion oncogenes is dependent on deregulation of ERBB-activated pathways and that a combination of RET and EGFR drugs could be more effective in treating RET fusion-positive tumors. Moreover, amplification of EGFR is a potential driver of resistance to cabozantinib in RET-rearranged lung cancers.

Abstract B30: Establishment and maintenance of a PDX Core Facility

Patient Derived Xenografts (PDX) are a powerful technology with the potential to unlock key mechanisms of disease for a variety of difficult to treat cancers and cancers with no applicable laboratory models. These models have a wealth of potential applications in tumor genetics, biomarker discovery, progression of metastasis, fate of circulating tumor cells, and in drug development of novel therapies for advanced and drug-resistant tumors. While several individual laboratories can excel at establishing such models, there are challenges, both procedural and technical, which present serious obstacles for established investigators as well as new investigators without prior experience with PDX models. Therefore, for some Institutions it is preferable to designate a centralized “PDX Core” to streamline the workflow that is required for the success of such program.

Here we describe and analyze the key components of a successful Academic PDX program. The PDX coordination branch carries out administrative duties to guarantee that the multiple components involved in generating PDXs are seamlessly integrated. In particular, it deals with logistical tasks including regulatory requirements (IRB, patient consent forms, IACUC), and sample accrual (patient selection, coordination with surgery schedules, specimen collection). In addition, it takes care of database maintenance and interface with clinical applications. The PDX laboratory branch meanwhile devises and implements standard operating procedures for sample processing, storage and implantation, and model maintenance, PDX characterization and distribution, and takes care of training of specialized personnel.

The establishment of such centralized PDX core at MSKCC resulted in a tenfold increase in the Institution's ability to generate PDX models within a year of inception of the program. In particular, for ongoing projects, we observed an increase not just in sample accrual but also in tumor take rate. In addition, the number of different tumor types implanted expanded substantially, from six to over twenty.

PDXs have the potential to be invaluable tools in understanding cancer and developing novel treatments. In our experience, the establishment of a centralized PDX core is both the most cost effective and efficient way for academic investigators to generate and gain access to novel, well characterized, and clinically annotated PDX models.

Citation Format: Marissa Mattar, Rajesh K. Uthamanthil, Elisa de Stanchina. Establishment and maintenance of a PDX Core Facility. [abstract]. In: Proceedings of the AACR Special Conference: Patient-Derived Cancer Models: Present and Future Applications from Basic Science to the Clinic; Feb 11-14, 2016; New Orleans, LA. Philadelphia (PA): AACR; Clin Cancer Res 2016;22(16_Suppl):Abstract nr B30.

Timing of CSF-1/CSF-1R signaling blockade is critical to improving responses to CTLA-4 based immunotherapy

Colony stimulating factor-1 (CSF-1) is produced by a variety of cancers and recruits myeloid cells that suppress antitumor immunity, including myeloid-derived suppressor cells (MDSCs.) Here, we show that both CSF-1 and its receptor (CSF-1R) are frequently expressed in tumors from cancer patients, and that this expression correlates with tumor-infiltration of MDSCs. Furthermore, we demonstrate that these tumor-infiltrating MDSCs are highly immunosuppressive but can be reprogrammed toward an antitumor phenotype in vitro upon CSF-1/CSF-1R signaling blockade. Supporting these findings, we show that inhibition of CSF-1/CSF-1R signaling using an anti-CSF-1R antibody can regulate both the number and the function of MDSCs in murine tumors in vivo. We further find that treatment with anti-CSF-1R antibody induces antitumor T-cell responses and tumor regression in multiple tumor models when combined with CTLA-4 blockade therapy. However, this occurs only when administered after or concurrent with CTLA-4 blockade, indicating that timing of each therapeutic intervention is critical for optimal antitumor responses. Importantly, MDSCs present within murine tumors after CTLA-4 blockade showed increased expression of CSF-1R and were capable of suppressing T cell proliferation, and CSF-1/CSF-1R expression in the human tumors was not reduced after treatment with CTLA-4 blockade immunotherapy. Taken together, our findings suggest that CSF-1R-expressing MDSCs can be targeted to modulate the tumor microenvironment and that timing of CSF-1/CSF-1R signaling blockade is critical to improving responses to checkpoint based immunotherapy.

SIGNIFICANCE

Infiltration by immunosuppressive myeloid cells contributes to tumor immune escape and can render patients resistant or less responsive to therapeutic intervention with checkpoint blocking antibodies. Our data demonstrate that blocking CSF-1/CSF-1R signaling using a monoclonal antibody directed to CSF-1R can regulate both the number and function of tumor-infiltrating immunosuppressive myeloid cells. In addition, our findings suggest that reprogramming myeloid responses may be a key in effectively enhancing cancer immunotherapy, offering several new potential combination therapies for future clinical testing. More importantly for clinical trial design, the timing of these interventions is critical to achieving improved tumor protection.

Genotype × environment interaction for long-yearling weight in Canchim cattle quantified by reaction norm analysis

The objective of this study was to investigate the presence of genotype × environment interactions (G×E) for long-yearling weight in Canchim cattle (5/8 Charolais + 3/8 zebu) in Brazil using reaction norms (RN). The hierarchical RN model included the fixed effect of age of the animal (linear coefficient) and random effects of contemporary groups and additive animal genetic intercept and slope of the RN and contemporary group effects as random effects. Contemporary groups as the most elemental representation of management conditions in beef cattle were chosen to represent the environmental covariate of the RN. The deviance information criteria demonstrated that a homoskedastic residual RN model provided a better data fit compared with a heteroskedastic counterpart and with a traditional animal model, which had the worst fit. The environmental gradient for long-yearling weight based on contemporary group effects ranged from -105 to 150 kg. The additive direct variance and heritability estimates increased with increasing environmental gradient from 74.33 ± 22.32 to 1,922.59 ± 258.99 kg(2) and from 0.08 ± 0.02 to 0.68 ± 0.03, respectively. The high genetic correlation (0.90 ± 0.03) between the intercept and the slope of the RN shows that animals with the greatest breeding values best responded to environmental improvement, characterizing scale effect as the source of G×E for long-yearling weight. The phenotypic plasticity demonstrated by the slope of the RN of the animal indicates the possibility to change genotype expression along the environmental gradient through selection. The results demonstrate the importance of accounting for G×E in the genetic evaluation of this population.

Appearances of paranasal fungal sinusitis on computed tomography

The primary invasive granulomatous form of fungal sinusitis, due to inhalation of aspergillus spores, is commonest in the Sudan and the Gulf states. This condition often presents clinically as a chronic, severe sinusitis which has not responded to antibiotics. On CT scanning, the major feature is a soft tissue mass, which is either homogeneous or has lower attenuation components. There may be erosion or expansion of the bony margins of the sinuses. Intraorbital and/or intracranial extension sometimes occur.

Degenerative changes and detection of plasminogen in fetal membranes that rupture prematurely

Amniochorion obtained at caesarean section and vaginal delivery, at 34-36 weeks gestation and at term, have been studied by electron microscopy and immunofluorescence for evidence of amniotic epithelial degenerative changes and the presence of plasminogen. When delivery was by caesarean section between 34-36 weeks, electron microscopy revealed no degenerative changes in four cases, but in two cases they were widespread. All but one membrane obtained at term showed only minimal amniotic epithelial cell degenerative changes, but extensive change was seen when delivery was premature after premature membrane rupture. Plasminogen was seen in amniotic epithelium proportional to the degree of cell degeneration; it was absent from healthy membranes. These findings demonstrate that degenerative changes are extensively present in membranes that rupture prematurely, particularly before the onset of premature labour, and suggest a role for plasminogen in membrane rupture.