An atlas of epithelial cell states and plasticity in lung adenocarcinoma

Understanding the cellular processes that underlie early lung adenocarcinoma (LUAD) development is needed to devise intervention strategies1. Here we studied 246,102 single epithelial cells from 16 early-stage LUADs and 47 matched normal lung samples. Epithelial cells comprised diverse normal and cancer cell states, and diversity among cancer cells was strongly linked to LUAD-specific oncogenic drivers. KRAS mutant cancer cells showed distinct transcriptional features, reduced differentiation and low levels of aneuploidy. Non-malignant areas surrounding human LUAD samples were enriched with alveolar intermediate cells that displayed elevated KRT8 expression (termed KRT8+ alveolar intermediate cells (KACs) here), reduced differentiation, increased plasticity and driver KRAS mutations. Expression profiles of KACs were enriched in lung precancer cells and in LUAD cells and signified poor survival. In mice exposed to tobacco carcinogen, KACs emerged before lung tumours and persisted for months after cessation of carcinogen exposure. Moreover, they acquired Kras mutations and conveyed sensitivity to targeted KRAS inhibition in KAC-enriched organoids derived from alveolar type 2 (AT2) cells. Last, lineage-labelling of AT2 cells or KRT8+ cells following carcinogen exposure showed that KACs are possible intermediates in AT2-to-tumour cell transformation. This study provides new insights into epithelial cell states at the root of LUAD development, and such states could harbour potential targets for prevention or intervention.

Prognostic value of comorbidities in older patients with cancer: the ELCAPA cohort study

BACKGROUND

In older patients, comorbidities competed with cancer for mortality risk. We assessed the prognostic value of comorbidities in older patients with cancer.

PATIENTS AND METHODS

We analysed all patients >70 years of age with colorectal, breast, prostate, or lung cancer included in the prospective ELCAPA cohort. The Cumulative Illness Rating Scale-Geriatrics (CIRS-G) score was used to assess comorbidities. The primary endpoint was overall survival (OS) at 3, 12, and 36 months. The adjusted difference in the restricted mean survival time (RMST) was used to assess the strength of the relationship between comorbidities and survival.

RESULTS

Of the 1551 patients included (median age 82 years; interquartile range 78-86 years), 502 (32%), 575 (38%), 283 (18%), and 191 (12%) had colorectal, breast, prostate, and lung cancer, respectively, and 50% had metastatic disease. Hypertension, kidney failure, and cognitive impairment were the most common comorbidities (67%, 38%, and 29% of the patients, respectively). A CIRS-G score >17, two or more severe comorbidities, more than seven comorbidities, heart failure, and cognitive impairment were independently associated with shorter OS. The greatest effect size was observed for CIRS-G >17 (versus CIRS-G <11): at 36 months, the adjusted differences in the RMST (95% confidence interval) were -6.0 months (-9.3 to -2.6 months) for colorectal cancer, -9.1 months (-13.2 to -4.9 months) for breast cancer, -8.3 months (-12.8 to -3.9 months) for prostate cancer, and -5.5 months (-9.9 to -1.1 months) for lung cancer (P < 0.05 for all).

CONCLUSIONS

Comorbidities' type, number, and severity were independently associated with shorter OS. A 17-point cut-off over 56 for the total CIRS-G score could be considered in clinical practice.

Abstract 113: An atlas of epithelial cell states and plasticity in lung adenocarcinoma

Understanding cellular processes underlying early lung adenocarcinoma (LUAD) development is needed to devise intervention strategies. While most if not all single-cell RNA sequencing (scRNA-seq) studies of lung cancer provided details on immune and stromal states, little insight is drawn to epithelial cells given their paucity (~4%) when performing unbiased scRNA-seq analysis without prior enrichment. Here, we performed in-depth scRNA-seq of enriched (by sorting for EPCAM+) epithelial cell subsets from 16 early-stage LUADs and 47 matching normal lung (NL) tissues. We also studied tissues from the same LUADs and adjacent NL by whole exome sequencing and a subset by high-resolution spatial protein and transcriptome analysis. We also performed scRNA-seq analysis of murine lungs from a human-relevant model of LUAD development following exposure to tobacco carcinogen, including strains with an alveolar type II (AT2) cell-specific lineage reporter. After extensive quality control, we retained 246,102 high quality human epithelial cells which comprised diverse normal alveolar and airway lineages as well as cancer cell populations. Diversity among cancer cells was strongly linked to LUAD oncogenic drivers. KRAS-mutant cancer cells were unique in their transcriptional features, strikingly reduced differentiation, low levels of copy number changes, and increased variability amongst the cells themselves. The local epithelial niche of LUADs, relative to that of NL, was enriched with intermediary cells in lung alveolar differentiation. A subset of these cells displayed elevated KRT8 expression (KRT8+ alveolar cells; KACs), increased plasticity and frequency of KRASG12D mutations, and its gene expression profiles were enriched in lung precancer and LUAD and signified poor survival. Notably, KACs harboring KRAS mutations were only found in the ecosystem of KRAS-mutant LUADs. Murine KACs were evident in lungs of tobacco carcinogen-exposed mice that develop KRAS-mutant LUADs but not in the saline-treated control group. While murine KACs emerged prior to tumor onset, they persisted for months after carcinogen cessation, and like their human counterparts, acquired driver Kras mutations, were poorly differentiated, and harbored KRAS-specific transcriptional programs. Spatial transcriptomics analysis showed that KAC and KRAS signatures were elevated in both murine and human tumors as well as in KACs that were in the local spatial vicinity of the LUADs. Organoids derived from lungs of tumor-bearing reporter mice were markedly enriched with KACs and were conspicuously sensitive to targeted inhibition of KRAS-G12D. This study provides new insights into the landscape of normal epithelial and malignant cells in LUAD, the role of alveolar intermediate subsets in development of the malignancy, particularly that driven by mutant KRAS, and, thus, potential targets for early interception.

Citation Format: Guangchun Han, Ansam Sinjab, Warapen Treekitkarnmongkol, Zahraa Rahal, Yuejiang Liu, Alejandra G. Serrano, Jiping Feng, Ke Liang, Khaja Khan, Wei Lu, Sharia Hernandez, Xuanye Cao, Enyu Dai, Yunhe Liu, Guangsheng Pei, Jian Hu, Lorena I. Gomez Bolanos, Edwin R. Parra, Tina Cascone, Boris Sepesi, Seyed Javad Moghaddam, Paul Scheet, Marcelo V. Negrao, John V. Heymach, Mingyao Li, Jichao Chen, Steven M. Dubinett, Junya Fujimoto, Luisa M. Solis, Ignacio I. Wistuba, Christopher S. Stevenson, Avrum Spira, Linghua Wang, Humam Kadara. An atlas of epithelial cell states and plasticity in lung adenocarcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 113.

Abstract 4460: Spatial profiling of immune biomarkers in resected treatment-naïve early stage lung adenocarcinoma

Background: Recently, neoadjuvant immunotherapy plus chemotherapy has been approved for treatment of resectable non-small cell lung carcinoma (NSCLC). Defining the immune landscape of these tumors and its spatial distribution will help to understand lung cancer biology. Here, we analyzed the distribution of immune-related biomarkers in tumor-defined regions and its associations with clinicopathological variables in resected lung adenocarcinomas using high-plex profiling approaches.

Methodology: Thirty-three FFPE tumor tissues from surgically resected treatment-naïve lung adenocarcinoma stage I/II were used to construct a tissue microarray from the MD Anderson ICON cohort. We used three 1-mm core per patient [2 from central tumor (CT), and 1 from invasive margin (IM)] and performed the GeoMx Digital Spatial Profiling protein protocol to assess 49 immune biomarkers. Pancytokeratin (panCK; epithelial), CD45 (immune) and SYTO 13 (nuclear) were utilized as morphology biomarkers. Regions of interests were placed in cores containing tumor, and segmented in ‘”Tumor (Tu)” (PanCK+) and the “tumor microenvironment (TME)” (PanCK-). Digital counts were normalized using background correction. Statistical analysis was performed using linear mixed model. A p value equal or less than 0.05 was considered significant.

Results: We first compared the relative counts of immune biomarkers in the TME in CT and IM. IM had higher CD3, CD8, CD45RO, as well as CD163 and STING (P ranges 0.006 to 0.035), while CT had higher PD1 (P 0.025). Then we analyzed differential biomarker expression by sex and smoking status. Females had higher counts of Immune related biomarkers: CD45, CD3, CD20, immune checkpoints: PD-L1, VISTA, CTLA4, LAG3, and ICOS, and myeloid: CD68, CD11c, CD163, and B2M (P ranges 0.0005 to 0.046). Smokers had higher counts of CD66b (P 0.007) and B2M (P 0.039) while never smokers had higher counts of HLA-DR, CD34, FoxP3, OX40L, Tim-3, and B7-H3 (P ranges 0.001 to 0.049). Finally, we analyzed Tu segments. IM had higher CD66b, VISTA, CD163, OX40L, HLA-DR, GZMB, STING, and CD8 (P ranges 0.007 to 0.045) than CT. Female patients had higher CD45, CD68, CD11c and CD163 (P ranges 0.008 to 0.045), and males had higher SMA (P 0.005). Smokers had higher CD66b (P 0.007), B2-microglobulin (P 0.043), and never smokers had higher HLA-DR, STING, CD34, CD44, FoxP3 and CD25 (P ranges 0.006 to 0.043).

Conclusions: In this study, biomarker analysis of treatment-naïve adenocarcinoma in CT areas and IM indicates a higher immune response in the IM and presence of inhibitory signaling inside the tumor. Our data also showed that tumors from females have higher immune response than tumors from males, which is concordant with previous studies. Distinct profiling by smoking status was also observed. Further analysis of gene expression analysis of this set is ongoing.

Citation Format: Sharia D. Hernandez, Wei Lu, Alejandra G. Serrano, Claudio J. Arrechedera, Beatriz Sanchez-Espiridion, Nejla Ozirmak, Max Molina, Larisa Kostousov, Sean Barnes, Khaja Khan, Ximing Tang, Junya Fujimoto, Edwin R. Parra, Gabriela Raso, Stephanie T. Schmidt, Carmen Behrens, John Heymach, Jianjun Zhang, Ken Chen, Boris Sepesi, Tina Cascone, Don Gibbons, Ignacio I. Wistuba, Cara Haymaker, Luisa M. Solis. Spatial profiling of immune biomarkers in resected treatment-naïve early stage lung adenocarcinoma. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 4460.

Evaluation of immune exhaustion across the evolution of oligometastatic clear cell renal cell carcinoma using spatially resolved tissue immune profiling: Secondary analysis of a prospective trial

701

Background: In patients with clear cell renal cell carcinoma (ccRCC), progressive local dysfunction in the tumor immune microenvironment (TIM) has been shown to play a significant role in clinical outcomes and response to therapy. Patients with oligometastatic ccRCC may particularly benefit from localized treatment such as radiation therapy. However, immune cell dysfunction in TIM and its influence on patient outcomes in oligometastatic ccRCC has yet to be defined. To address this unmet need we conducted spatially resolved immune profiling on longitudinally collected patient samples from our ongoing prospective phase II trial (NCT03575611) of radiation therapy in oligometastatic ccRCC. This methodology offers a unique multidimensional evaluation of TIM in relation to histological architecture. Methods: Thirty-seven FFPE tumor samples from 32 patients with ccRCC (primary, 21; oligometastatic, 16) were obtained to construct a tissue microarray (TMA) and assess tumor immune activation biomarkers using the nanoString GeoMx Digital Spatial Profiler (DSP). We used the pan-cytokeratin, CD3, CD68, and SYTO13 as morphology biomarkers to profile, in each TMA core, a panel of 49 protein immune biomarkers in each of these compartments (DSP areas of illumination/AOI): tumor (panCK+), T-cell (CD3+), and macrophage (CD68+). In addition, DSP Immunofluorescence images were used to calculate cell density, and cell-to-cell distances of T cell and Macrophage. A median of 7 AOIs (range 3-9) were assessed in each tissue. Comparisons were made between primary and metastatic tissues by Mann-Whitney test with correction false discovery rate of 1%. Regression models were used to evaluate association between protein counts within each compartment and clinical outcomes. Results: In T-cells, CD127 was significantly higher in primary ccRCC tissue (p=0.00009) while FOXP3 and LAG3 were significantly increased in oligometastatic tissue (p=0.001). Macrophages in oligometastatic tissue expressed higher CD163 and ARG1 (p=0.01 and 0.02, respectively), compared to those in primary tissue. Average distances between macrophage and the nearest T cell were significantly shorter in oligometastatic tumor tissue compared to those in primary ccRCC (22.9 vs 28.8 µm, p=0.04). Higher PD-1 (R2=0.26) with lower CD127 (R2=0.22) and CD27 (R2=0.21) expression in T cells in primary tumor predicted earlier time to first diagnosis of metastases after nephrectomy (all p<0.05). Conclusions: Spatially resolved immune profiling revealed that the TIM in metastatic tumor harbored more T cells exhibiting markers of suppression with closer spatial interaction to macrophages compared with primary tumors. Furthermore, TIM with T cell exhaustion in primary tumors predicted earlier diagnosis of metastasis after nephrectomy.

The Role of Nerve Fibers in the Tumor Immune Microenvironment of Solid Tumors

The importance of neurons and nerve fibers in the tumor microenvironment (TME) of solid tumors is now acknowledged after being unexplored for a long time; this is possible due to the development of new technologies that allow in situ characterization of the TME. Recent studies have shown that the density and types of nerves that innervate tumors can predict a patient's clinical outcome and drive several processes of tumor biology. Nowadays, several efforts in cancer research and neuroscience are taking place to elucidate the mechanisms that drive tumor-associated innervation and nerve-tumor and nerve-immune interaction. Assessment of neurons and nerves within the context of the TME can be performed in situ, in tumor tissue, using several pathology-based strategies that utilize histochemical and immunohistochemistry principles, hi-plex technologies, and computational pathology approaches to identify measurable histopathological characteristics of nerves. These features include the number and type of tumor associated nerves, topographical location and microenvironment of neural invasion of malignant cells, and investigation of neuro-related biomarker expression in nerves, tumor cells, and cells of the TME. A deeper understanding of these complex interactions and the impact of nerves in tumor biology will guide the design of better strategies for targeted therapy in clinical trials.

Quantitation of Pulmonary Surfactant Protein SP-B in the Absence or Presence of Phospholipids by Enzyme-Linked Immunosorbent Assay

We have developed an enzyme-linked immunosorbent assay (ELISA) that uses polyclonal or monoclonal anti-surfactant protein SP-B antibodies to quantitate purified SP-B in chloroform/methanol and in chloroform/methanol extracts of whole pulmonary surfactant at nanogram levels. This method has been used to explore the effect of the presence of different phospholipids on the immunoreactivity of SP-B. Both polyclonal and monoclonal antibodies produced reproducible ELISA calibration curves for methanolic SP-B solutions with protein concentrations in the range of 20-1000 ng/mL. At these protein concentrations, neither dipalmitoylphosphatidylcholine, dipalmitoylphosphatidylglycerol, nor phosphatidylcholine or phosphatidylglycerol from egg yolk had significant effects on the binding of antibodies to SP-B up to protein-to-lipid weight ratios of 1:20. Coating of ELISA plates with SP-B concentrations higher than 1 microg/mL produced a substantial decrease in the binding of antibodies to the protein that was prevented by the presence of negatively charged but not zwitterionic phospholipids. Characterization of the secondary structure of SP-B by far-UV circular dichroism showed that phospholipids induced pronounced changes on the conformation of SP-B when the solvent was evaporated and dry lipid-protein films were formed, a necessary step to expose protein to antibodies in ELISA. Under these conditions, negatively charged lipids, but not zwitterionic ones, induced a marked decrease on the ellipticity of SP-B that would be associated with a conformation that is significantly more exposed to antibodies.

Microstructure and dynamic surface properties of surfactant protein SP-B/dipalmitoylphosphatidylcholine interfacial films spread from lipid-protein bilayers

Suspensions of dipalmitoylphosphatidylcholine (DPPC) bilayers containing 5, 10 or 20% (w/w) surfactant protein SP-B have been reconstituted and spread at air-liquid interfaces. Compression isotherms of DPPC/SP-B monolayers spread from these preparations were qualitatively comparable to the isotherms of the corresponding DPPC/SP-B monolayers spread from solvents. SP-B was squeezed-out at higher pressures from vesicle-spread films than from solvent-spread monolayers. SP-B caused a marked decrease on the rate of relaxation of DPPC collapse phases to equilibrium pressures in all the lipid/protein films assayed. This stabilizing effect was higher in vesicle-spread than in solvent-spread monolayers. Inclusion in the films of traces of the fluorescent probe NBD-PC (1 mol%) and use of a fluorescent derivative of SP-B labeled with a rhodamine derivative, Texas Red, allowed for direct observation of protein and lipid domains at the interface by epifluorescence microscopy. Upon compression, SP-B altered the packing of phospholipids in the bilayer-spread films, observed as a SP-B-induced reduction of the area of liquid-condensed domains, in a way similar to its effect in solvent-spread monolayers. SP-B was not associated with condensed regions of the films. Fluorescence images from vesicle-spread films showed discrete fluorescent aggregates that could be consistent with the existence of lipid-protein vesicles in close association with the monolayer. Both the retention of SP-B at higher surface pressures and the greater stability of collapse phases of DPPC/SP-B films prepared by spreading from liposomes in comparison to those spread from solvents can be interpreted as a consequence of formation of complex bilayer-monolayer interacting systems.