Patent Application
20240288449
The present invention relates to the field of anticancer immunotherapy. In particular, the present invention concerns the role of the MAdCAM-1/α4β7 axis the gut microbiota in the efficacy of cancer treatments and provides methods for determining if a patient is likely to benefit from a cancer treatment, more precisely, an immuno-oncology (I-O) therapy, such as a treatment comprising administration of an antibody directed against immune checkpoint blockers PD1, PD-L1 or PD-L2 alone or together with CTLA4. The present invention also provides methods to improve the efficacy of such a treatment in patients in need thereof.
During the last decade, the gut microbiome has garnered much attention in the context of evolving immuno-oncology with compelling evidence of the role of intestinal dysbiosis in primary resistance to immune checkpoint blockade across a large range of advanced malignancies. This contention was supported by numerous epidemiological studies pointing out the deleterious impact of broad-spectrum antibiotics in the clinical outcome of anti-PD-1/anti-PD L-1 antibodies (Abs) in stage III and IV lung, kidney, bladder cancers and melanoma. Of note, decreased richness of the intestinal ecosystem was associated with a tumor microenvironment poor in T cell infiltrates.
The increased awareness of the potential contribution of gut dysbiosis in treatment failure led many investigators to describe metagenomics-based intestinal blueprints associated with resistance to immunotherapy. However, to which extent cancer-associated dysbiosis precedes the tumorigenic process and therefore is causatively linked to neoplasia development or is merely its direct consequence remains unknown.
β7 integrin-containing heterodimers play a central role in leukocyte homing to the gut and retention at epithelial surfaces. The β7 integrin subunit can form a heterodimer with the α4 (CD49d) integrin subunit resulting in the α4β7 integrin (also called “Lymphocyte Peyer's patch Adhesion Molecule-1” (LPAM-1)). Through interaction with its counter-receptor Mucosal Addressin Cellular Adhesion Molecule-1 (MAdCAM-1), α4β7 integrin mediates lymphocyte adhesion and diapedesis from the circulation across the vascular endothelial barrier into gut-associated secondary lymphoid tissue (GALT) or the intestinal lamina propria (LP). The transmembrane adhesion molecule MAdCAM-1 is constitutively expressed on LP venules and GALT high endothelial venules (HEV) and inducible by proinflammatory cytokines (Briskin et al., 1997; Gorfu et al., 2009; Ogawa et al., 2005). Retinoic acid (RA) expressing CD103+ dendritic cells govern the expression of α4β7 integrin and CCR9 receptors in the GALT to enable binding to MAdCAM-1 and homing of lymphocytes to the gut mucosa through the intestinal CCR9 ligand, CCL25. By preventing the influx and extravasation of inflammatory β7+ T cells from the circulation to the gut, antibodies targeting α4β7 or MAdCAM-1 significantly reduced the severity of colitis in animal models and in patients suffering from inflammatory bowel disease (Hassan-Zahraee et al., 2018; Reinisch et al., 2021). There is a growing awareness of the role played by intestinal microbiota in maintaining the regulatory functions of intestinal Treg and TH17 cells to ensure gut epithelial barrier integrity (Littman and Rudensky, 2010; Pandiyan et al., 2019).
Beyond their seminal homeostatic role in the intestines, TH17 also control extraintestinal inflammatory lesions (Krebs et al., 2016; Lee et al., 2011a; Magnuson et al., 2015; Morton et al., 2014; Wu et al., 2010). Importantly, TH17 and their lineage-related FoxP3+ regulatory cells blunt antitumor immunosurveillance by restoring tolerance during inflammation induced in the course of chronic carcinogenesis (Cochaud et al., 2013; Fridman et al., 2012; Guéry and Hugues, 2015; Langowski et al., 2006; Young, 2016). RORγt expression identified a phenotypically stable population of tumor-infiltrating Tregs in humans and mice that are induced by gut commensals (Sefik et al., 2015). Conditional RORγt knockout mice showed reduced tumor incidence or polyp formation, through an indirect mechanism involving dendritic cells (Blatner et al., 2012; Rizzo et al., 2018).
Immune checkpoint inhibitors (ICI) reinstate tumor immunosurveillance and have become the standard of care for the clinical management of several histological types of cancers (Brahmer et al., 2015; Robert et al., 2011). Primary resistance to ICI has been mainly attributed to low tumor mutational burden and poor intrinsic antigenicity of tumor cells (Riaz et al., 2016; Rizvi et al., 2015), defective antigen presentation (Spranger et al., 2015), intratumoral T cell exhaustion (Smyth et al., 2016), genomic defects in the Interferon-γ (IFNγ) signaling pathway (Gao et al., 2016), CSF1-dependent tumor associated macrophages (Neubert et al., 2018), and immunosuppressive metabolic cues (Smyth et al., 2016). In addition, several studies indicated that a deviated gut microbiome repertoire negatively impact ICI efficacy, suggesting the crucial importance of commensals in affecting the cancer-immune set point (Chen and Mellman, 2017). Retrospective and prospective studies analyzing the impact of antibiotics (ATB) on patients survival revealed their negative predictive impact on clinical outcome during immune checkpoint blockade (Elkrief et al., 2019b; Mohiuddin et al., 2021; Routy et al., 2017; Tinsley et al., 2019; Derosa et al. 2022). Meta-analyses revealed that ATB uptake is more harmful on clinical outcome when administered prior to the first systemic administration of (rather than during) anti-PD1/PDL-1 Abs, suggesting that recolonization post-ATB course rather than direct iatrogenic effects may be deleterious (Derosa et al., 2021). Indeed, ATB-treating patients tended to be colonized by distinct species (such as Hungatella hathewayi (Derosa et al. 2022). These epidemiological findings led to the hypothesis that ATB, by affecting the composition of the intestinal ecosystem, could tilt the equilibrium between intratumoral effector and regulatory T cells by affecting the gut/tumor trafficking of intestinal homeostatic regulatory T cell subsets (38), thereby aggravating cancer immunosuppression.
Recently, the inventors conducted meta-analyses which revealed that ATB uptake is more harmful on clinical outcome when administered prior to (rather than during) the first systemic administration of anti-PD1/PDL-1 Abs, suggesting that recolonization post-ATB course rather than direct iatrogenic effects may be deleterious (Derosa et al., Cancer Discovery, 2021).
In the experimental part which follows, the inventors confirm this hypothesis and show the pathways through which the gut microbiota distantly affects the evolution of several types of cancers and the patient's response to immunotherapy.
Indeed, the inventors showed that ATB-induced dysbiosis (Routy et al., 2018; Vétizou et al., 2015) results in the disruption of the interaction between Mucosal Addressin Cell Adhesion Molecule-1 (MAdCAM-1) and its receptor α4β7 expressed by enterotropic T cells.
When the course of ATB is stopped, selected species from the Enterocloster gen.nov. of bacteria (such as Enterocloster clostridioformis comb.nov (Haas and Blanchard, 2020), previously classified in the Clostridia class of bacteria) take over and downregulate ileal Mucosal Addressin Cell Adhesion Molecule-1 (madcam1) gene and cell surface protein expression on ileal lamina propria and high endothelial venules. Loss of MAdCAM-1 provokes the exodus of the gut homing α4β7+ TH17 and RORγt+ Treg CD4+ T cells to the tumor beds, tolerizing the cancer microenvironment and culminating in resistance to PD-1 blockade.
The present invention is based on these results and first pertains to a method for in vitro assessing whether an individual having a cancer also has a cancer- or antibiotics-associated dysbiosis, based on down regulation of MAdCAM in serum or ileum and/or Gut OncoMicrobiome Signature (GOMS). This method can also be used to follow-up the effects of a microbiota-centered intervention (MCI) in a patient.
Diagnosing such a cancer- and/or ATB-related dysbiosis indicates that the subject needs a compensatory microbiota-centered intervention (MCI), especially before receiving a treatment with an immune checkpoint inhibitor (ICI) or with another immuno-oncology (I-O) therapy.
The use of such MCI, including oral vancomycin antibiotics, phages and rare-cutting endonucleases able to kill bacteria of the Enterocloster gen. nov. clade, Akkermansia spp and/or Akkermansia muciniphila, possibly mixed with other beneficial bacteria, retinoic acid, fecal microbial transplantation, and mixtures thereof, for restoring sensitivity to an I-O therapy, is also part of the present invention.
The invention also pertains to combined therapies for cancer patients having a low level of serum soluble MAdCAM, with (i) an anti-PD1 or an anti-PDL1 antibody and (ii) an anti-IL-17A or an anti-IL-17R antibody and/or a recombinant IL-7.
The present invention also relates to screening methods for identifying agents capable of normalizing MAdCAM-1 expression levels in endothelial cells of the lamina propria.
A-D. Relative transcription (A) and protein (B-D) levels of Madcam1 gene product obtained with RT-PCR (A), immunohistochemistry staining of MAdCAM-1 in LP venules (upper micrograph pictures) and high endothelial venules (HEV, lower micrograph pictures) (B), flow cytometry gating on CD45− cells of LP (C) or ELISA (D) of ileal (A-D) or colonic (A) tissues in C57BL/6J mice after continuous broad-spectrum antibiotics (ATB: Ampicillin, Colistin, Streptomycin) or after a 4 to 7 days spontaneous recolonization (ATBRECO) following ATB stop. Each dot represents one ileum or colon. E-F. Quantitative RT-PCR of relative madcam1 gene expression in ileal mucosae of mice treated with 8 days of oral vancomycine (E) or broad spectrum ATB followed by 4-7 days recolonization (ATBRECO) (F), and then supplemented with gavage with Enterocloster clostridioformis once prior to sacrifice 7 days later. Each graph depicts a representative experiment (comprising 5-6 mice/group) out of 2-3 yielding similar results. Panel A or E left panel pooled the data of 2 experiments. G. Idem as in E. but oral gavage was performed in C57BL/6J mice reared in SPF conditions (no ATB conditioning) using bacteria spp. aligned in the x axis. H. Left panel: Impact of fecal microbiota transplantation (FMT) in 3 days—ATB treated recipient mice using feces from NSCLC patients at diagnosis (prior to PD-1 blockade) on ileal Madcam1 gene expression normalized on the naïve specific-pathogen free (SPF) mice group (log 10 axis). Each experiment used a different FMT donor and comprised 5-6 animals/group. Each dot represent on ileum. Right panel: non supervised hierarchical clustering of the taxonomic composition of donor feces (defined using shot gun MG sequencing), selecting bacteria of high prevalence >25% and clinically relevant. I. Same experiments as in A-D performing flow cytometric analysis of ileal lamina propria CD4+ T cell subsets (α4β7+ vs CD25+ FoxP3+ Treg vs RORγt+ CD4+ (TH17) cells) during continuous ATB or at the ATBRECO phase, each dot representing one ileum. A representative Facs analysis is depicted out of 2 yielding similar results. Also refer to
A-B. Flow cytometric determination of photoconverted (A) or CFSE labeled (B) gut-originating cells in secondary lymphoid organs (mesenteric LN, spleen, tdLN: tumor draining LN) assessed in Kaede mice after illumination according to the schema (A, left panel) or by CFSE injection into the mLN in wild-type mice (B, left panel). Left panels depict the graphical schema of the experimental setting and right panels show the relative accumulation (log 2 Fold Change (FC) of photoconverted (PC) versus resident (non photoconverted (NPC) cells in each subset detailed in lines using a color gradient (A) or CFSE labeled vs resident (non labeled) cells (B) 24 hours after tissue UV-A illumination or CFSE injection in the mLN, in various secondary lymphoid organs. Statistics: test of Mann-Whitney with p-value adjusted by Benjamini-Hochberg method, *p<0.05, **p<0.01, ***p<0.001. C. Effect of the neutralization of MAdCAM-1 using αMAdCAM-1 mAb on the recirculation of gut-derived cells towards the tdLN in Kaede (left) or in mLN-CFSE injected WT mice (middle panel) or towards a controlatreal LN (right panel). Each dot represents one mouse. A representative experiment out of two is depicted, reaching similar conclusions. D. Volcano plot depicting the differential gene transcription in RNA sequencing of α4β7high versus α4β7neg CD4+ lymphocytes cell-sorted from mLN in 5 MCA205 tumor-bearing animals. Volcano plots were generated computing for each gene product expressed in each cell type for 5 mice i) the log 2 of fold ratio (FR) among the mean relative abundances of transcripts after normalization in α4β7high versus α4β7neg CD4+ lymphocytes (x axis); ii) the co-log 10 of p-values deriving from Mann-Whitney U test calculated on relative abundances in absolute values (y axis). Black and grey dots are considered significant (p<0.05) while back dots are not (p>0.05). E. Experimental setting (left panel) and flow cytometric evaluation of the IL-17A+ secreting α4β7+ Treg (middle panel) or α4β7+ Tconv (right panel) CD4+ T cells within CFSE+ (originating from the mLN, grey dots) or CFSE-cells (tdLN resident cells, black dots) reaching the tdLN during the ATBRECO phase following ATB conditioning (Ampicillin, Colistin, Streptomycin). Each graph depicts two independent and pooled experiments (comprising 5-6 mice/group). F. Flow cytometric determination of CFSE+ Treg cell reaching the tdLN or the contralateral LN in MCA205 bearing WT versus Madcam1−/− mice (not treated with ATB) after CFSE injection in mLN. Each dot represents one mouse, a representative experiment out of two yielding similar results being depicted. G. Id. As in E. but in Kaede mice treated or not with anti-MADCAM-1 Abs where flow cytometry could identify Treg using membrane stainings for CD25 and CD127. H. Experimental setting to assess the short term (4-7 days) effects of bacterial recolonization after stopping the ATB course onto recirculation of enterotropic CD4+ T cells, with or without enforced oral gavage with Enterocloster spp. (left panel). Flow cytometric phenotyping of TH17 cells expressing high levels of CD25 (Tr7-like, middle panel) or TH17 conv. (right panel) in Kaede mice within PC (originating from the mLN) or NPC cells (tdLN resident cells) reaching the tdLN during the ATBRECO phase (G). Each dot represents one mouse and graphs depicts two pooled experiments. ANOVA statistical analyses (Kruskal-wallis test) or Wilcoxon matched-pairs signed rank test: raw p values are indicated. I. Single cell transcriptomics analyses of CFSE+ mLN-derived CD4+ T cells emigrating to tdLN in MCA205 tumor bearers treated with water (group 1), with ATB 7 days, ATB being continued (group 2), or ATB being stopped since 4 days (ATBRECO) (group 3) or replaced by oral gavage with E. clostridioformis for 4 days (group 4), the experiment containing 5 animals/group. UMAP gene patterns in 4 subsets of emigrating CFSE+ CD4+ T cells in plate-based full-length single-cell RNA-seq data by unsupervised clustering, overlaying the four groups of mice (I, left panel). Volcano plots of differential gene expression patterns associated with each specific cell type according to Log 2 p values and fold ratios for the proportion of each subtype compared with the 3 others (Treg (I, middle panel) and proliferating cells (I, right panel). Refer to
A. Tumor growth kinetics of subcutaneous MCA205 (syngeneic of C57BL/6J) implanted in wild type (wt) versus Itgb7 or Madcam1 gene deficient mice. Means±SEM of tumor sizes among 5-6 mice/group overtime in 2 treatment groups (anti-PD1 versus isoCtl Ab). B-D. Tumor growth kinetics or tumor luminescence after implantation of subcutaneous MCA205 (B), mammary 4T1 (syngeneic of BALB/c, C) and lung-orthotopic TC1-luc (syngeneic of C57BL/6J, D) tumors in animals treated with isotype control, anti-α4β7 mAb or −αMAdCAM-1 mAb while receiving anti-PD1 therapeutic antibodies (or iso-Ctl Abs). For lung-orthotopic TC1 tumor, a whole-body luminescence using IVIS, Imaging System was used to quantify luciferase-expressing TC1 kinetics in lungs. Ratios between pre- and post-PD-1 blockade with isotype control mAb, αα4β7 mAb or αMAdCAM-1 mAb were calculated. E. Flow cytometric analysis of α4β7 expression on CD4+ T cell splenocytes and tumor infiltrating lymphocytes (TILs) in wild type (wt) versus Madcam1 gene deficient mice. Each dot represents one mouse. The graph pooled 2 independent experiments of 5 mice/group. Student's t test. F. Non supervised hierarchical clustering represented in a heatmap showing the relative percentages of various immune α4β7+CD4+ T cell subsets (determined by flow cytometric analyses of TILs dissociated from established s.c. MCA205) in mice treated with neutralizing αMAdCAM-1 mAb or isotype control mAb. G-I. Intracellular flow cytometric analysis of Rorγt expression in α4β7+ Treg TIL in wild type (wt) versus Madcam1 gene deficient mice (G) or in wt MCA205 tumor bearing mice receiving ATB conditioning regimen ((4 days post-ATB stop (ATBRECO-S) or 12 days post ATB stop (ATBRECO-L)) treated or not with anti-PD1 Abs (H,I). Each dot represents one mouse. A representative experiment out of 2 independent ones yielding similar results and comprising 5 mice/group is depicted (G, I). Student's t test (G). J. Flow cytometric analyses of tumor infiltrating cells (TILs) for CCR5+CD8+ effector T cells in s.c. MCA205 tumor-bearing mice treated with isotype control mAb, αα4β7 mAb or αMAdCAM-1 mAb in the setting of PD-1 blockade. K. Schema of the experimental setting (left panel) and cross sectional study of tumor sizes (means±SEM) of s.c. 4T1 WT (left) or 4T1 IL-22Rα1 KO (right) tumor cell lines at sacrifice after a therapy with αPD-1 mAb during the ATBRECO phase and systemic neutralization of IL-17A using αIL-17A mAb (ip) in parallel to the immunotherapy. In each experiment, four αPD-1 intraperitoneal (i.p.) injections were administered every three days during the ATBRECO phase. Each experiment comprised 6 mice/group. A representative experiment out of 2-3 is depicted. ANOVA statistical analyses (Kruskal-wallis test): *raw p values are indicated.
A. Spearman correlation between RT-PCR-determined ileal Madcam1 gene expression levels and serum soluble MADCAM-1 (determined by ELISA) in 65 MCA205 tumor bearing mice, each dot representing one animal. B. ELISA monitoring of serum levels of soluble MADCAM-1 in 301 NSCLC patients belonging to 2 independent cohorts (C01, C02, left panel) according to history of their recent ATB uptake (right panel) compared with serum levels in healthy volunteers (HV, n=70). Each dot represents one patient's serum. C-D. Alpha and beta diversity of the taxonomic content of the intestinal microbiota according to the serum sMADCAM-1 levels in NSCLC patients. MGS diversity-based richness evaluation in shot gun MG sequencing according to the median of sMADCAM-1 in the whole population composed of 95 NSCLC patients (C) and interindividual variability of the composition in both groups (D). E-F. Kaplan Meier survival curves and Cox regression linear analysis of the predictive value of serum levels of sMADCAM-1 split according to the median of each cohort in two independent (left and right columns) cohorts of stage IV NSCLC treated with αPD-1 mAb immunotherapy (refer to Tables 5-6 for patient description and multivariable analysis) for PFS (E) and OS (F). G. Supervised hierarchical clustering of MGS (with prevalence>10%) according to sMADCAM-1 serum levels (<(Low) or >(High) median), the color code featuring the relative abundance of each significant species, aligned in bar graph in the left column using various pipelines and algorithms. H. Relative abundance of E. clostridioformis in percentage among all MGS species in fecal materials of 95 patients, split in two groups according to their sMADCAM-1 serum levels (<(Low) or >(High) median). Each dot represents one feces/serum/patient. Student t′ test.
A. Heatmap of log 2 Fold Change ratios between ATB-treated and not treated (Water) ileum, colon, as well as tumor beds for chemokines (left panel) and cytokines/transcription factors (right panel) expression profiles in ELISA (left panel) and RT-PCR (right panel) of the tissue lysates. Statistics: test of Mann-Whitney with p-value adjusted by Benjamini-Hochberg method, *p<0.05, **p<0.01, ***p<0.001. B. Evaluation of CD31 expression on ileal HEV by IHC in ATB-treated mice. C. RT-PCR evaluation of the impact of 7 days conditioning with various antibiotic regimens or after ATBRECO on relative transcription levels of Madcam1 gene in ileal tissue (left) or in mesenteric lymph nodes (right). Continuous (c.), Ampic: c. Ampicillin, Colistc: c. colistin, Streptoc: c. Streptomycin, Vancoc: c. vancomycin, Erythrc: c. Erythromycin. Each dot represents one ileum. The graph gathered 2-3 experiments comprising 5 mice/group. D-E. RT-PCR assessment of the relative expression of Madcam1(D, E) and VCAM1 (E) genes in Peyer's patches (PP) (D) or various lymph nodes (LN) (mLN: mesenteric LN, sk: skin LN, td: tumor draining LN) (E). Each dot represents one ileum. The graph gathered 2 experiments comprising 5 mice/group. F. Culturomics, in aerobic and anaerobic conditions, of ileal material harvested from mice treated with various ATB regimen (refer to C) and sacrificed at day 4 after their stop. G. Qualitative identification by mass spectrometry of cultivated species in each condition. H-I. RT-PCR evaluation of murine Foxp3, RORc and IL17a genes in ileal tissue of mice treated or not with ATB (H) and Spearman correlations between ileal Madcam1 and Rorc or Foxp3 expression levels in mice (1). Each dot represents one animal. Concatenated data from 3 independent experiments. J. Spearman correlations between RT-PCR-based transcriptional levels of human FOXP3 (left), and IL-17A (right) and Madcam-1 genes in intestinal biopsies collected during endoscopic intervention in 9 control (ATB free) patients and 7 ATB-treated patients (Table 1). Each dot represents one biopsy either from ileum, colon and cecum, a single patient being represented 1 to 3 times. ANOVA statistical analyses (Kruskal-wallis test): *p<0.05, **p<0.01, ***p<0.001.
A. Schematic overview of the experimental setup of UV-A illumination of the intestine (ileum, caecum, mLN) in tumor-bearing Kaede mice and flow cytometric gating strategy of Photoconverted cells (PC) in tdLN (upper panels) and mLN (lower panels). PC frequencies in target organs before, 5 minutes and 24 hours after laparotomy and UV-A illumination of the intestine compartment are indicated. B. Id. As in A. with detailed percentages of PC cells in mLN, spleen, tdLN and tumor beds at 24 hours, each dot representing one mouse. A typical experiment out of 3 yielding similar results is depicted. C. Experimental setting (left panel) and flow cytometric determination of CFSE+ cells 24 hrs after CFSE injection in mLN in different tissues (x axis), each dot representing one mouse. Data from 3 independent experiments are pooled. D. Id. As in B. monitoring the mobilization of enterotropic cells after UV-A illumination of “intestine” (ileum+caecum+mLN) versus ileum only, towards tdLN or tumor beds, each dot representing one mouse. A typical experiment out of 2 yielding similar results is depicted. E-F. Flow cytometric evaluation of gut integrin α4β7 expression by resident (NPC) or recirculating (PC) CD4+ T cells in the spleen, tdLN or tumor bed of Kaede mice according to the area of UV-A illumination (ileum vs “intestine” (ileum+caecum+mLN) or in WT mice CFSE injected in mLN (F). G-H. Flow cytometric evaluation of different CD4+ Treg cell types (varying in the expression of α4β7, IL-17, IL-22) in tdLN traced by CFSE labeling in animals treated or not with ATB, and bearing subcutaneous MCA205 and subjected to CFSE injection in the mLN 24 hours before. ANOVA statistical analyses (Kruskal-Wallis test): *annotated raw p values.
A. Experimental design for—scRNA seq data collection. Single cell transcriptomics was performed on CFSE+ CD4+ T cells harvested from tdLN 24 hrs post-CFSE inoculation in mLN in 4 experimental groups of mice (water, continuous ATB, ATB RECO, E. clostridioformis) containing 5 MCA205 tumor-bearing animals. B. Absolute numbers of CFSE+ cells recovered in tdLN for each condition after quality control and sequenced. C. Volcano plot depicting the differential gene transcription in RNA sequencing of CFSE+ CD4+ T cells belonging to the Treg subset cells in the group ATBRECO versus the water group (control). Volcano plots were generated computing for each gene product expressed in each cell type for 5 mice i) the log 2 of fold ratio (FR) among the mean relative abundances of transcripts after normalization in group 3 versus group 1 (x axis); ii) the co-log 10 of p-values deriving from Mann-Whitney U test calculated on relative abundances in absolute values (y axis). Blue (downregulated gene products) and red (upregulated gene products) dots are considered significant (p<0.05) while back dots are not (p>0.05). D. Idem as C but comparing the gene transcripts in Treg versus the other T cell subsets within the ATBRECO (group 3, left panel) or ATB E. clostridioformis (group 4). E. Idem as D but comparing the gene transcripts in proliferative T cell subset versus the other T cell subsets within the negative control group (water, group 1, left panel) and the ATBRECO (group 3, right panel). F. Idem as E but comparing the gene transcripts in cells belonging to cluster 1 (in UMAP,
A. Spearman correlations between ileal Madcam1 and tumor size in MCA205 tumor bearing (anti-PD1 or isotype Ctl Ab-treated) mice. Each dot represents one mouse. ANOVA statistical analyses (Kruskal-wallis test): *annotated raw p values. B. Flow cytometric assessment of RORγt+ Treg (Tr17) among α4β7+ or α4β7− CD4+ TIL fractions in subcutaneous MCA205 tumors. Each dot represents one tumor and mouse. C. Experimental setting of PD1 blockade during ATB intervals of various durations in tumor bearers. D-F. Flow cytometric assessment of RORγt+ or IL-17A+ IL-22+ cells within Treg in the mLN (D) or in subcutaneous MCA205 (E) or 4T1 (F) tumors. G. Flow cytometric determination of IL-17A+ IL-22+ secreting Treg (left) and α4β7+ Treg cells (right) in subcutaneous (s.c.) MCA205 tumors treated by PD1 blockade after ATB stop and spontaneous ATBRECOL+/−enforced gut colonization with E. clostridioformis or L. reuteri bacteria during αPD-1-based therapy. Each dot represents the tumor size in each animal at sacrifice, in a typical experiment containing 6 mice/group out of two yielding similar results. Refer to Table 4 for detailed phenotypes. ANOVA statistical analyses (Kruskal-Wallis test): *annotated raw p values.
A. Tumor sizes at sacrifice in MCA205 tumor bearing mice treated with 3 days ATB and gut humanized using FMT from PD1 refractory NSCLC patients (FMT NR), supplemented or not with oral gavage of live or pasteurized A. muciniphila SGB9228 (A. muc, left panel). Flow cytometric evaluation of α4β7 and/or CCR9 expression by CD4+ T cells in the tdLN draining s.c. MCA205 sarcoma (middle and right panels). B. Effects of anti-MADCAM-1 neutralizing antibodies on the compensatory effects of pasteurized A. muciniphila SGB9228 for the response to PD1 blockade depicted as a ratio between tumor sizes in anti-PD1-treated FMT NR avatar mice compensated with the bacterium+/−anti-MADCAM-1 Abs. C-D. Flow cytometric evaluation of IL-17A and IL-22 cytokine secretion by CD4+ T and Treg cells in the tdLN of mice colonized with FMT NR and supplemented by oral gavage with Akkermansia muciniphila (A. muc), receiving or not the αMAdCAM-1 mAb treatment and treated with αPD-1 mAb (C). Effects of the anti-MADCAM-1 neutralizing antibodies on the preventive effects of pasteurized A. muciniphila SGB9228 to prevent the gut exodus of Tr17 depicted as a ratio between IL-17+IL-22+ Treg or total CD4+ T cells reaching the tdLN in anti-PD1-treated FMT NR avatar mice compensated with the bacterium+/−anti-MADCAM-1 Abs (D). ANOVA statistical analyses (Kruskal-wallis test) or Wilcoxon matched-pairs signed rank test: *raw p values are annotated. E. Graphical abstract illustrating the recirculation toward the tumor microenvironment of gut-primed IL-17A+ IL-22+ RORγt+α4β7+ Treg (Tr17) cells occurring when there is a disruption of MADCAM-1 expression in ileal HEV in a context of ATB-induced dysbiosis.
A. Cox linear regression analysis evaluating the Hazard Ratio (HR) for the risk of progression or death during PD1 blockade and serum sMADCAM-1 (monitored in ELISA) in each cohort C01 and C02 described in Table 5, using restricted cubic splines (RCS) with 4 knots. The number of knots was chosen according to the Akaike Information Criterion. Left panel (C01): N=115 NSCLC patients, 62 deaths, median OS=15 months (95% C.I: 11.2-25 mo), HR=0.92 (95% C.I: 0.88-0.97, p=0.001) Right panel (C02): N=186 NSCLC patients, 83 deaths, median OS=12 months (95% C.I: 9-17 mo), HR=0.97 (95% C.I: 0.94-0.99, p=0.02). B. Prevalung study: Serum sMADCAM-1 levels in CVD smokers before lung cancer incidence in those patients who will be diagnosed with lung cancers (N=9) versus not (N=56 controls) within 2 years of follow up. C. Spearman correlation between serm sMADCAM-1 and circulating α4ρ7+ Rorγt Treg. D-E. Non supervised hierarchical clustering of the taxonomic composition based on shot gun MG sequencing of feces in NSCLC patients according to sMADCAM-1 paired serum levels and ANOVA statistical analysis of the segregation between taxonomic MGS-based clustering and median of sMADCAM-1.
A. Concentrations of sMAdCAM_1 in serum of patients at baseline prior to Durvalumab and during therapy at 1 month and 4 months in responders (elite patients, PFS>5 months) and non responders (progressors OS>6 months), each dot representing one time point and patient, each patient being represented 3 times. B-C. Log rank univariate analysis of the predictive value of sMAdCAM_1 and Kaplan Meier curves of survival (OD, B) and PFS (C) in patients segregated according to the median value of MAdCAM_1 within the whole cohort.
According to a first aspect, the present invention pertains to a method for in vitro diagnosing cancer- or antibiotics-associated intestinal dysbiosis in an individual having a cancer, comprising measuring soluble MAdCAM-1 in a serum sample from said individual (e.g., by ELISA), wherein a reduced level of serum soluble MAdCAM-1 indicates that the individual has a cancer- or antibiotics-associated intestinal dysbiosis.
Alternatively, the above method can be performed by assessing the expression of MAdCAM-1 in ileal lamina propria (LP) venules or in high endothelial venules (HEV) in an ileal biopsy from the patient, wherein a reduction of MAdCAM-1 expression in LP or HEV indicates that the individual has a cancer-associated ileopathy/dysbiosis. In such a case, the expression of MAdCAM-1 can be measured by RT-PCR or with an ELISA or flow cytometry or immunohistochemistry performed on the biopsy.
The present invention thus relates to the use of serum soluble MAdCAM-1 level, as a marker of cancer- or antibiotics-associated dysbiosis, wherein reduced level of serum soluble MAdCAM-1 is a marker of cancer- or antibiotics-associated dysbiosis.
As shown in the experimental part below, the inventors have demonstrated that the level of soluble MAdCAM-1 in the serum of an individual also is a marker of resistance or sensitivity to an immuno-oncology (I-O) therapy, wherein a reduced level of serum soluble MAdCAM-1 is a marker of resistance to the I-O therapy.
According to the present invention, the level of soluble MAdCAM-1 in the serum of an individual can thus be used to assess whether the individual is likely to resist to an immuno-oncology (I-O) therapy (low level). A normal (or elevated) level of serum soluble MAdCAM-1 is a marker of clinical benefit.
In the present text, the phrase “I-O therapy” includes immune checkpoint inhibitors (ICI), as well as CAR-T cells, adoptive TIL transfer and combinations thereof. In the context of the present invention, “I-O therapies” also include combined therapies including one of the above I-O agents and other antineoplastic treatments, such as chemotherapy, especially any combination of an immune checkpoint inhibitor (ICI) with taxanes, permetrexed, cis-platin and/or oxaliplatinum, or EGFR inhibitors.
In the context of the present invention, “ICI” include anti-PD1 antibodies (Ab), anti PDL-1 Ab, anti-CTLA4 Ab, anti-Lag3 Ab, anti-Tim3 Ab, anti-TIGIT Ab, anti-OX40 Ab, anti-41 BB Ab, anti-VISTA Ab, bispecific antibodies targeting PD1 and Lag3, as well as other molecules exerting the same function(s), such as non-Ab molecules blocking any of the above immune checkpoints. According to a particular embodiment, the I-O therapy includes an anti-PD1/PDL-1 Ab, e.g. monoclonal Ab blocking PD1 or PDL-1.
In certain cases of treatment resistance, it has been shown that ICI drugs can not only be useless, but even have deleterious effects, leading to rapid tumor progression (i.e., hyperprogressive disease or HPD). Identifying a patient likely to resist to an I-O treatment is thus of major importance to decide not to administer the I-O treatment to this patient, at least not without a compensatory treatment or combined therapy to avoid HPD onset.
The present invention is particularly useful for assessing the resistance status of an individual who has taken broad spectrum antibiotics during a period ranging from 60 days before and 42 days after the first administration of an I-O therapy/ICI. Indeed, as shown in the experimental part below, broad spectrum antibiotics increase the risk of dysbiosis associated with I-O resistance.
When performing the above method, the level of serum soluble MAdCAM-1 is considered as reduced when it is inferior to a predetermined threshold. This threshold can be, for example, the median of the levels of soluble MAdCAM-1 in a representative cohort, such as a cohort of individuals suffering from cancer, preferably from the same cancer as the tested individual. Even more preferably, the representative cohort is a cohort of individuals who suffer from the same cancer as the tested individual and are receiving the I-O therapy/ICI in a same therapeutic scheme (especially, as a same line of therapy, at least differentiating 1 L vs 2 L therapy). The skilled person can refine the threshold by measuring the median of the levels of soluble MAdCAM-1 in a cohort of patients who share the same therapeutic history and received the same I-O therapy. In the experimental results disclosed below, the value of this median was respectively, 177.1 ng/ml and 233.3 ng/ml in a discovery and a validation cohort in advanced non small cell lung cancer (NSCLC) patients treated with anti-PD1/L-1 antibodies with or without chemotherapy (Example 1), while it was 88.8 ng/ml in a cohort of patients with kidney cancer treated with nivolumab in second line therapy (Example 2.1), and of 158.8 ng/ml in a cohort of cancer patients with pre-treated metastatic bladder treated with durvalumab (Example 2.2).
Alternatively, the predetermined threshold can be calculated from a cohort of individuals not suffering from cancer. For example, the level of serum soluble MAdCAM-1 can be considered as reduced if it is in the inferior tertile of a cohort of individuals not suffering from cancer.
The above method is particularly interesting for patients suffering from a cancer amenable to immunotherapy alone or combined with chemotherapy or tyrosine kinase inhibitors or hormonotherapy (androgen or oestrogen deprivation, or LHRH antagonists).
Examples of such cancers include breast cancer, Chronic Myelomonocytic Leukemia (CMML), colorectal cancer, kidney cancer, lung cancer (e.g. NSCLC), urothelial cancer, melanoma, ovary cancer, stomach and oesophageal cancers, mesothelioma, hepatocarcinoma, prostate cancer and any mismatch repair insufficiency (MSI) high tumor for which there is an agnostic histology with FDA-approval for anti-PD1 Abs. According to a particular embodiment, the individual suffers from a cancer selected from the group consisting of NSCLC, melanoma, breast cancer, kidney cancer, bladder cancer and colorectal cancer.
The present invention also relates to a theranostic method for determining if an individual having a cancer needs a compensatory microbiota-centered intervention (MCI) before administration of an immune-oncology (I-O) therapy, comprising assessing, by the above method, whether the individual has a cancer- or antibiotics-associated dysbiosis, wherein if the individual has a cancer- or antibiotics-associated dysbiosis, he/she needs a MCI before administration of the treatment with an I-O therapy.
In certain cases, such as lung cancers, the MCI can be combined to another treatment, for example to chemotherapy for avoiding HPD.
In the present text, the phrase “microbiota-centered intervention (MCI)” designates any treatment having a direct or indirect effect on the intestinal microbiota composition.
Examples of MCI according to the invention include:
In the above, phages and endonucleases are considered as “killing bacteria of the Enterocloster gen. nov. clade” if they are able to kill bacteria of at least one of the species of the Enterocloster list corresponding to clading taxonomy reported in Haas and Blanchard, Int J Syst Evol Microbiol 2020; 70:23-34.
This list includes:
The present invention also pertains to the use of an agent for MCI, as those listed above, for treating cancer in combination with an I-O therapy in an individual having a cancer-induced dysbiosis.
Another object of the present invention is a method for following-up the success of a MCI, comprising measuring the level of serum soluble MAdCAM-1, wherein a normalized level of serum soluble MAdCAM-1 indicates that the MCI successfully restored gut eubiosis, or at least partially corrected the dysbiosis.
When performing this method, the level of serum soluble MAdCAM-1 can be considered as “normalized” if it is higher than a predetermined threshold (as described above). Alternatively, it can be considered as “normalized” or at least “partially normalized” if it is higher than before the MCI.
The present invention also relates to the use of serum soluble MAdCAM-1 level, as a biomarker predicting early NSCLC incidence in high risk heavy smokers experiencing a cardiovascular event (HRHSCV), wherein reduced level of serum soluble MAdCAM-1 is a marker of predicting the onset of NSCLC.
In the experimental part which follows, the inventors demonstrated that in the setting of intestinal dysbiosis, PD1 blockade facilitates the amplification and homing to tumors of TH17/Tr17 in the GALT. They also showed that in such a situation, blocking IL-17A bioactivity could circumvent the deleterious effects of dysbiosis.
Hence, combining IL-17A and PD1 blockades is clinically very interesting for patients who have been identified as likely to resist to a treatment with an anti-PD1 Ab by one of the methods described above.
Besides, IL-7 controls α4β7 integrin expression and imprints gut-homing specificity on T cells, acting on CD127 highly expressed on emigrating Tr17 in inflammatory bowel disease (Belarif et al. J. Cin Invest May 2019, vol 129, numero 5).
As a result, combining recombinant IL-7 with PD1 blockade is another option clinically very interesting for patients who have been identified as likely to resist to a treatment with an anti-PD1 Ab by one of the methods described above.
Another aspect of the present invention is thus the use of a combined therapy comprising (i) an anti-PD1 or an anti-PDL1 antibody and (ii) an anti-IL17A or anti-IL-17R antibody and/or recombinant IL-7, for the treatment of cancer in a patient having a low serum level of soluble MAdCAM.
In the above combined therapy, the two or more agents can be administered together or separately, starting by blocking IL-17R or IL-17 and then inhibiting PD1 or PDL-1/PDL-2.
Other aspects of the present invention relate to methods and tools for identifying agents capable of normalizing MAdCAM-1 expression levels in endothelial cells of the lamina propria.
Such a method comprises (i) in vitro contacting cells expressing MAdCAM-1, or engineered to express a reporter gene under the control of the MAdCAM-1 promoter, with compounds to be tested and (ii) assessing which of said compounds induce the expression of MAdCAM-1 or of the reporter gene expressed under the control of the MAdCAM-1 promoter.
Examples of cells that can be used to perform this method include Human umbilical vein endothelial cells (HUVECs), transformed sinusoidal endothelial cells (TSECs), bEnd.3 cells and derivatives thereof (in particular, engineered cells derived from these cell lines by stably integrating a reporter gene under the control of the MAdCAM-1 regulatory elements).
According to another method, the compounds to be tested are administered to gut humanized avatar mice, and the expression of MAdCAM-1 in endothelial cells of said mice is measured.
The two methods can also be combined.
In such a combined screening method, the in vitro results are completed by a step of validation in avatar mice, wherein avatar mice are gut humanized mice that were pre-treated with 8-10 days broad spectrum ATB and then recolonized by a human FMT product by oral gavage, wherein:
A method according to the invention can also comprise:
A screening platform for performing the above method is also part of the present invention. Such a platform comprises cells expressing a reporter gene under the control of the MAdCAM-1 promoter (e.g., HUVECs, TSECs or bEnd.3 cells, or cells derived therefrom) and a robot configured for picking agents in a bank of products, delivering these to cultivated cells and assessing expression of the reporter gene (for example by measuring fluorescence if the reporter gene expresses a fluorescent protein).
As mentioned above, a low level of soluble MAdCAM in the serum of a cancer patient is a marker of gut dysbiosis in this patient, likely to induce resistance to I-O therapy, so that this patient needs a compensatory preparation before the I-O therapy, and can advantageously benefit from a treatment comprising the steps of:
Measuring the level of soluble MAdCAM in the serum of a cancer patient before beginning or continuing an I-O therapy (for example, by the above-described method) helps determining if the patient is in need of a combined treatment (MCI+I-O).
Patients for whom this measure is particularly useful are those who have taken broad spectrum antibiotics during a period ranging from 60 days before and 42 days after the first administration of an ICI, especially an anti-PD1/PDL-1 Ab, because they are more likely to have a severe gut dysbiosis.
Advantageously, a new measure of the serum level of soluble MAdCAM is done before step (ii) above to check whether the MCI has reduced the cancer-associated dybiosis.
When performing the above combined treatment (MCI+I-O), an individual who has received broad spectrum antibiotics less than 60 days ago can advantageously receive a fecal microbial transplantation, possibly enriched with pasteurized Akkermansia spp and/or Akkermansia muciniphila; and an individual who has not received broad spectrum antibiotics less than 60 days ago can advantageously receive live Akkermansia spp and/or Akkermansia muciniphila.
Other characteristics of the invention will also become apparent in the course of the description which follows of the biological assays which have been performed in the framework of the invention and which provide it with the required experimental support, without limiting its scope.
Fecal material. For feces collection, ancillary studies were conducted at Institut Gustave Roussy/France according to the ethical guidelines and approval of the local CCPPRB. The study name was “Oncobiotics”, B2M ethics protocol number PP: 15-013. Written informed consent in accordance with the Declaration of Helsinki was obtained from all patients. For collection of endoscopic and blood samples a clinical study “Einfluss von Antibiotika auf das Darm-Chemokinnetzwerk bei Patienten mit soliden Tumoren” was conducted at University Clinics Heidelberg/Germany according to the ethical guidelines and approval of the Regierungsprasisium Karlsruhe. Eligible patients had stage IIIA-IV non-small-cell lung cancer (NSCLC) with squamous or non-squamous histology and had documented recurrence or progression after at least one prior line of treatment. Participant sites were Gustave Roussy Cancer Campus (Villejuif, France). NSCLC patients received anti-PD-1 mAb, nivolumab, as part of the treatment modalities approved by the European Medical Agency (EMA). Tumor response was assessed using the Response Evaluation Criteria in Solid Tumors version 1.1 (RECIST1.1). Computer tomography (CT) scans were performed at baseline and every 8 to 12 weeks for the first year and then every 12 to 15 weeks until disease progression. Feces were collected according to International Human Microbiome Standards (IHMS) guidelines (SOP 03 V1) before the first injection (TO). In brief, a collection kit including an anaerobic generator (Biomerieux) was given to patients. Samples were collected by patients at home, and frozen 4 to 24 h later at −80° C. at Gustave Roussy Cancer Campus in plastic tubes (Plastic vessel by 1000-Sarstedt) with or without BHI+2% glycerol.
Collection of endoscopic samples and blood samples. Eligible patients underwent ileo-colonoscopy according to clinical standard protocols for non-study related indications (Table 1) between July 2018 and November 2019. When feasible, endoscopic biopsy of the mucous membrane of the terminal ileum, cecum, and right and left colon was performed on each patient. Tissue samples were either snap frozen in liquid nitrogen and stored at −80° C. or immersed into 2% PFA for histology. In addition, two blood samples (10 ml EDTA tubes) were collected before ile-colonoscopy. All included patients responded a questionnaire to assess dietary history and clinical baseline data was retrieved from the local clinical information system.
Cell culture, reagents and tumor cell lines. MCA-205 fibrosarcoma cells, MC38 and RET melanoma cells (i.e. melanomas generated by transgene-enforced expression of the Ret proto-oncogene under the control of the metallothionein-1 promoter driving spontaneous melanomagenesis, kindly provided by Prof. Viktor Umansky) (all syngeneic for C57BL/6 mice) were cultured at 37° C. in the presence of 5% CO2 in RPMI 1640 containing 10% FCS, 2 mM L-glutamine, 100 UI/ml penicillin/streptomycin, 1 mM sodium pyruvate and MEM non-essential amino acids (henceforth referred to as complete RPMI 1640). Luciferase-transfected TC-1 cell lines (syngeneic for C57bl6 mice, kindly provided by Pr. Eric Deutsch, Institut Gustave Roussy, France) were cultured at 37° C. in the presence of 5% CO2 in complete RPMI 1640 and 1 mM Hepes buffer. Cell lines were regularly tested for mycoplasma contamination and were not used after more than 10 passages.
Mice. All animal experiments were carried out in compliance with French and European laws and regulations. The local institutional animal ethics board and French Ministére de la Recherche approved all mouse experiments (permission numbers: 2016-049-4646, 2017_049_99741, 2019_036_21124). Experiments were performed in accordance with Government and institutional guidelines and regulations. Female C57BL/6 were purchased from Harlan (France). Mice were used between 7 and 12 weeks of age. MAdCAM-1-KO and ITGB7-KO mice were a kind gift from Angela Schippers, University hospital Aachen, Aachen, Germany. MAdCAM-1-KO and ITGB7-KO mice and control littermates were backcrossed on the C57BL/6 background and were obtained from an in-house-breed at the local animal care facility at the University Hospital Aachen. CCR5-KO mice were a kind gift from Christophe Combadiére, Hopital Salpetriére, Paris, France. CCR9-KO mice were a kind gift of Reinhold Forster, University hospital Hannover, Hannover, Germany. CCR5-KO and CCR9-KO mice were both maintained on a C57BL/6 background. All mouse experiments were performed at the animal facility in Gustave Roussy Cancer Campus where animals were housed in specific pathogen-free conditions.
Antibiotic treatments. If not indicated otherwise, mice were treated with a broad spectrum antibiotic solution (ATB) containing ampicillin (1 mg/ml), streptomycin (5 mg/ml), and colistin (1 mg/ml) (Sigma-Aldrich) added to the sterile drinking water of mice. In experiments where single antibiotics were used the concentrations were ampicillin (1 mg/ml), streptomycin (5 mg/ml), colistin (1 mg/ml), erythromycin (1 mg/ml) or ciprofloxacin (0.1 mg/ml) respectively. Solutions and bottles were changed 2 times weekly. Antibiotic activity, in the experiments where the antibiotic mix was used, was confirmed by cultivating fecal pellets resuspended in BHI+15% glycerol at 0.1 g/ml on COS (Columbia Agar with 5% Sheep Blood) plates for 48 h at 37° C. in aerobic and anaerobic conditions weekly. Duration of ATB treatment was slightly different (7-14 days before aPD-1 treatment) based on the experimental settings and is indicated in results for each experiment. In brief, to compromise the efficacy of anti-PD-1 mAb, mice were treated with ATB for one to two weeks before tumor implantation and ATB was continued throughout the experiment or discontinued on the day of anti PD-1 treatment initiation as indicated in the individual experiment. In the context of fecal microbial transplantation experiments, mice received 3 days of ATB before undergoing fecal microbial transplantation the next day by oral gavage using animal feeding needles. In the TC-1 model, ATB treatment was started 3 days before tumor injection and discontinued one day before the first bacterial gavage.
Subcutaneous model of MCA-205 sarcoma, MC38 and RET melanoma. Syngeneic C57BL/6 mice were respectively implanted with 0.8×106 MCA-205 sarcoma, 1.0×106 MC38 or 0.5×106 RET melanoma cells subcutaneously and treated intraperitoneally (i.p.) when tumors reached 20 to 40 mm2 in size with anti-PD-1 mAb (250 μg/mouse; clone RMP1-14,) or isotype control (clone 2A3). Mice were injected 4 times at 3-day intervals with anti-PD-1 mAb. Tumor length and width was routinely monitored every 3 times per week by means of a caliper. In experiments using anti-a4b7 mAb (DATK32, 200 μg per mouse) or anti-MadCAM mAb (MECA-367, 200 μg per mouse), mAbs (or their isotype controls, clone 2A3 in both cases) were injected i.p. every 3 days starting from day 0 until the final anti-PD-1 injection. All antibodies were purchased from BioXcell, NH, USA.
Orthotopic luciferase engineered-TC-1. C57BL/6 mice were anesthetized with isoflurane. Under sterile conditions a lateral incision was made on the chest wall of each mouse and 6×105 TC-1-Luc cells in 10 μl Matrigel (Corning) were injected into the lung. The skin incision was closed with a surgical skin clip. Tumor growth was monitored twice weekly on an IVIS Imaging System 50 Series (Caliper Life Sciences/Xenogen). Starting on day 3 mice were injected with anti-PD-1 mAb or recommended isotype according to the same dose and schedule as above.
FMT experiments. Fecal microbiota transfer (FMT) was performed by thawing fecal material. Mice were placed in a new cage. Two hundred μL of the suspension was then transferred by oral gavage into each germ-free or ATB pre-treated (3 days) recipient. In addition, another 100 μL was applied on the fur of each animal. Two weeks after FMT, tumor cells were injected subcutaneously or orthotopically and mice were treated with anti-PD-1 or isotype controls as mentioned above. Oral bacterial gavage was performed on the same days as aPD-1 treatment as described below.
Oral bacterial gavage with commensal species. Akkermansia muciniphila CSUR P2261, and A. indistinctus CSUR P723 were provided by the Institut hospitalo-universitaire Méditerranée Infection, Marseille, France. Enterococcus hirae 13144 isolates were originally isolated from spleens or mesenteric lymph nodes of SPF mice treated with CTX at Gustave Roussy Cancer Campus. A. muciniphila was grown on COS plates in an anaerobic atmosphere created using 3 anaerobic generators (Biomerieux) at 37° C. for at least 72 h. E. hirae 13144 were grown on 5% sheep blood enriched Columbia agar for 24 h at 37° C. in aerobic conditions. Colonization of ATB pre-treated or GF C57BL/6 mice was performed by oral gavage with 100 μl of suspension containing 1×108 bacteria. For bacterial gavage: suspensions of 101 CFU/mL were obtained using a fluorescence spectrophotometer (Eppendorf) at an optical density of 600 nm in PBS. Five bacterial gavages were performed for each mouse, the first 24 h before the first injection of antiPD-1 mAb and subsequently four times on the same day as anti-PD-1 mAb injection. The efficacy of colonization with E. hirae 13144 was confirmed by culturing the feces 48 h post gavage. Fecal pellets were harvested and resuspended in BHI+15% glycerol at 0.1 g/ml. Serial dilutions of feces were plated onto 5% sheep blood enriched Columbia agar and incubated for 48 h at 37° C. in aerobic and anaerobic conditions. After 48 h, single colonies were isolated and Gram staining was performed. The identification of specific bacteria was accomplished using a Matrix-Assisted Laser Desorption/Ionisation Time of Flight (MALDI-TOF) mass spectrometer (Andromas, Beckman Coulter, France).
Flow cytometry analyses. Tumors, mesenteric lymph nodes (mLN), draining lymph nodes (tdLN) and spleens were harvested at different time points as indicated in the individual experiments. Excised tumors were cut into small pieces and digested in RPMI medium containing Liberase™ at 25 μg/mL (Roche) and DNase1 at 150 UI/mL (Roche) for 30 minutes at 37° C. and then crushed and filtered twice using 100 and 70 μm cell strainers (Becton & Dickinson). Lymph nodes and spleen were crushed in RPMI medium and subsequently filtered twice through a 100 μm cell strainer. Four million tumor cells, lymph node cells or splenocytes were pre-incubated with purified anti-mouse CD16/CD32 (clone 93; eBioscience) for 30 minutes at 4° C., before membrane staining. For intracellular staining, the Foxp3 staining kit (eBioscience) was used. Dead cells were excluded using the Live/Dead Fixable Yellow dead cell stain kit (Life Technologies) or LIVE/DEAD™ Fixable Aqua Dead Cell Stain Kit. Anti-mouse antibodies for CD3 (145-2C11), CD4 (GK1.5 or RM4-5), CD8 (53-6.7), CD25 (3C7), CD44 (IM7), CD45 (30-F11), CD62L (MEL-14), CD127 (A7R34), Foxp3 (150D), RORγt (B2D), CXCR3 (FAB1685P), CXCR5 (J252D4), CCR2 (SA203G11), CCR4 (2G12), CCR5 (HM-CCR5), CCR6 (29-2L17), CCR7 (4B12), CCR9 (CW-1.2), beta7 (FIB504), a4b7/LPAM-1 (DATK32 and REA457), MadCAM-1 (MECA367) all purchased from Miltenyi, BioLegend and eBioscience) were used to stain cells. Stained samples were acquired on CytoFLEX S 13 colours (Beckman Coulter) or a BD Facs CANTO II (BD) and analyses were performed with Kaluza software 1.5 (Beckmann Coulter). T central memory (TCM) gating: after gating on CD3+ alive, CD4+ were selected, then TCM were identified as being either CD62L+. Effector memory T (TEM) cells were selected as being CD62L and CD44+. Th17 gating depended on the mouse model used. The Kaede fluorochrome does not retain its photoconverted state upon fixation with PFA. Th17 gating: after gating on CD3+ alive, CD4+ were selected, then Th17 were identified as RORγt+. For unfixed cells Th17 were identified as CXCR3− and CCR6+. Treg gating: after gating on CD3+ alive, CD4+ were selected, for fixed cells Treg were identified as FoxP3+CD25+. For unfixed cells Treg were identified as CD127− CD25+. For definition of cut off values Fluorescence minus one (FMO) controls were used as appropriate.
Immunohistochemistry. 3 μm-thick sections of formalin-fixed, paraffin embedded murine ileum and colon were prepared as “swiss rolls” and were mounted on poly-L-lysine-coated slides, deparaffinized and hydrated through graded alcohols to water. Antigen retrieval was carried out by heating sections for 30 minutes in buffer at 98° C., (respectively with 0.01 M sodium citrate buffer, pH 6.0, for the staining of CD3 and with 1 mM EDTA, pH 8.0 for the staining of FoxP3). Endogenous peroxidase activity was inhibited with 3% hydrogen peroxidase (DAKO) for 10 minutes and then saturated for 20 minutes with PowerVision IHC/ISH Super Blocking Solution (Leica Biosystems, #PV6122). Without washing, the rabbit anti-human CD3 polyclonal Ab (ready-to-use, DAKO, #IS503) and rabbit anti-mouse Foxp3 polyclonal Ab (2 μg/ml, Invitrogen, #PA1-46126) primary antibodies were applied and incubated for 1 hour, followed by the secondary Ab, PowerVision poly-horse radish peroxidase-conjugated anti-rabbit antibody (Leica Biosystems #PV6119). Peroxidase was detected by means of the Di Amino Benzidine (DAB)-peroxidase substrate kit (DAKO), and the sections were counterstained with Mayer's hematoxylin. The immunohistochemistry staining of CD4 was performed on an automated immunostainer (The BenchMark ULTRA, Ventana, IGR). Heat-induced antigen retrieval in EDTA buffer (pH 8.0) for 32 minutes at 95° C. was performed. The primary monoclonal anti-mouse CD4 antibody (1,246 μg/mL, #ab183685, Abcam) was diluted 1:500 in antibody diluent (Zytomed), and the slides were incubated for 1 hour at 37° C. The biotin-free peroxydase system of detection technique with DAB as chromogen was applied (kit ultraView Universal DAB Detection kit, Ventana). The slides were also counterstained by means of a hematoxylin kit (Ventana).
Kaede experiments. The Kaede mice were a kind gift from Michio Tomura, Kyoto University, Kyoto, Japan and were backcrossed and maintained on the C57BL/6 background. Kaede transgenic mice were anesthetized with 2 to 2.5% isoflurane and administered buprenorphine (0.01 mg/kg) i.p. for analgesia. For photoconversion of ilea abdominal skin and peritoneum were cut at the midline to access the intraperitoneal terminal ileum. For photoconversion of the colon, ileum or ileum including the mesenteric lymph nodes, the cecal pole was identified and the cecal pole including the terminal ileum, the mesenteric lymph nodes and the proximal colon was gently mobilized through the midabdominal incision onto a sterile plastic coated surgical drape. Non target structures were covered with aluminium foil. The ventral and dorsal parts of the target structures were exposed to violet light emitted from a 395 nm wave length emitting Diode (Winzwon) light for 30 seconds each. After illumination the tissue was moistened with sterile isotonic sodium chloride and gently repositioned into the peritoneal cavity. The peritoneal membrane was closed by continuous stitch with a 5-0 monofilic nylon suture (Ethicon). The skin was closed with two 9 mm wound clips (EZ Clip Kit).
Tracing migration of leukocytes from mesenteric and tumor draining lymph nodes using CFSE. C57B16 mice were anesthetized with 2 to 2.5% isoflurane and administered buprenorphine (0.01 mg/kg) i.p. for analgesia. Abdominal skin and peritoneum were cut at the midline to access the mesenteric lymph nodes. The mesenteric lymph nodes were gently mobilized through the midabdominal incision onto a sterile plastic coated surgical drape. Ileum draining mesenteric lymph nodes were visually identified according to their vasculature. The two most prominent mesenteric lymph nodes were injected with 100 μM CFSE diluted in 5 μl PBS using a 30G insulin syringe. After reposition of the mesenteric lymph node the peritoneal membrane was closed by continuous stitch with a 5-0 monofilic nylon suture (Ethicon). In the case of tumor draining lymph nodes the abdominal skin was incised in the midline sparing the peritoneum. The tumor draining lymph nodes were visualized by gentle separation of the abdominal skin and the peritoneum using scissors. The tumor draining lymph nodes were injected with 100 μM CFSE diluted in 5 μl PBS using a 30G insulin syringe. The skin was closed with two 9 mm wound clips (EZ Clip Kit).
RNA extraction and rtPCR. Lysis and extraction protocols were identical for human and mouse samples. Tumor or intestinal samples were snap frozen in liquid nitrogen in RLT Plus buffer containing 0.1% beta mercaptoethanol. On the day of extraction samples were thawed at 4° C. and homogenized on a microtube homogenizer (Benchmark Scientific) in RNA free glass bead tubes (Dutscher). Total RNA extraction and genomic DNA removal were performed with the RNeasy Mini kit (Qiagen), following the manufacturer's recommendations. A maximum of 1 μg of RNA, measured by using a NanoDrop™ Spectrophotometer (Thermo Fischer Scientific), was reverse transcribed into cDNA with a mix composed of SuperScript III Reverse Transcriptase (Life Technologies), RNaseOUT™ Recombinant Ribonuclease Inhibitor (Life Technologies), Random primers (Promega) and Deoxynucleoside Triphosphate Set, PCR grade (Roche Diagnostics).
Single cell RNA sequencing by Rhapsody. After isolation of CFSE+ CD4+ T cell by flow cytomatry, cells were washed in cold PBS 10.000 cells and loaded onto a BD Rhapsody™ cartridge and processed according to manufacturer's instructions for targeted single-cell RNA-seq using the predesigned Immune Response Panel (Mouse). The library was clustered at 1.75 μM on a NextSeq500 system (Illumina) to generate −40.000 paired end (2*75 bp) reads per cell using High Output v2 chemistry. Sequenced single-cell data was demultiplexed using bcl2fastq2 v2.20.
Quantitative gene expression assay. Expression of B2M, FoxP3, IFNγ, IL-10, IL-17, MadCAM, Ppia, RORc, TNFα (all from Life Technologies) was analyzed with the TaqMan® Gene Expression Assay using the Universal Master Mix II on a StepOnePlus™ Real-Time PCR System (Life Technologies). Amplifications were carried out using the following ramping profile: 1 cycle at 95° C. for 10 min, followed by 45 cycles of 95° C. for 30 s, 60° C. for 1 min. Quantitative RT-PCR data were normalized to the expression levels of the housekeeping genes P2M or Ppia, as indicated in each figure, by means of the 2-ΔCt method multiplied by 106.
Tissue lysis and chemokine analysis. Intestinal and tumoral samples were snap frozen in liquid nitrogen in a non-denaturing cell lysis buffer containing 50 mM Tris HCL pH 7.4, 150 mM NaCL, 300 mM Sucrose, 10 mM EDTA and 0.1% Triton 100X. For subsequent lysis samples were thawed at 4° C. and lysed on a tube homogenizer (Precellys) in ceramic beads lysis tubes (Precellys). Tissue homogenate was centrifuged at 4000 g for 5 min. The supernatant was used for subsequent analysis. Chemokine concentrations in the tissue lysate were determined according to manufacturer recommendation using CCL2, CCL3, CCL4, CCL5, CCL25 and MadCAM Duoset ELISA kits (RnD) or using Legendplex Mouse proinflammatory chemokine panel (Biolegend) with cytometric analysis performed on a CytoFLEX S (Beckmann coulter).
Statistics. Mice. Data analyses were performed either with the statistical environment R (http://www.R-project.org/) or Prism 6 (GraphPad, San Diego, CA, USA). Tumor size differences were calculated using non-parametric t-test. All reported tests are two-tailed and were considered significant at p<0.05. Patients. MADCAM-1 as a predictor of response to PD1 blockade in NSCLC patients. The prognostic role of sMadCAM1 level was studied using restricted cubic splines (RCS) with 4 knots. The number of knots was chosen according to the Akaike Information Criterion. We tested for non-linearity in the biomarker effect. When linearity hypothesis was not rejected, the linear coding was used to estimate the effect of the biomarker.
We reported that gut sterilization of mice with a broad-spectrum ATB cocktail (Ampicillin, Colistin and Streptomycin) blunted the anticancer efficacy of PD-1 blockade (Routy et al., 2018a). We first analyzed ATB-induced fluctuations of the expression of chemokines and integrin ligands in the intestines of MCA205 tumor-bearing mice treated with anti-PD-1 mAb. ATB induced a significant loss of expression of most ileal chemokines and Madcam1 gene products (
Fecal microbial transfer (FMT) of stools from melanoma patients who benefited from PD1 blockade could circumvent primary resistance to ICI in one third of metastatic melanoma recipients (Baruch et al., 2021; Davar et al., 2021). Indeed, distinct donor FMT failed to transfer clinical benefit to recipients (Table 1). Hence, we tested whether random FMT from lung or kidney cancer patients into avatar mice (as previously reported, (Routy et al., 2018a, 2018b) treated by ATB could downregulate ileal Madcam1 gene expression in recipient mice reared in SPF conditions. Three out of 6 FMT did so (
Oral supplementation of FMT recipients with various commensals, including the pro-inflammatory Atopobium parvulum, L. reuteri but not H. hathewayi could further influence Madcam1 ileal gene expression (
ATB-induced Madcam1 gene downregulation in the ileum paralleled and correlated with that of regulatory cytokines and transcription factors (such as 1117a, 1122, Foxp3, RORc) (
As shown in mice, we confirmed the coordinated inhibitory effects of ATB on the intestinal expression of Madcam1, Foxp3 and Rorc in 16 patients who took ATB and underwent intestinal endoscopy and biopsies for various indications (
Altogether, broad-spectrum ATB regimen induced a sustained downregulation of the ileal mucosal addressin MADCAM1 in mice and humans, that markedly correlated with reduced ileal Foxp3, 1117a and RORc expression.
We hypothesized that the loss of MADCAM-1 molecules could affect the homing or the retention of enterotropic T cells expressing its α4β7 receptor in the GALT. To track the migration of ileal T cells in tumor-bearing hosts, we utilized two complementary experimental strategies. First, we used a transgenic Kaede mouse model that allows to irreversibly label cells without disrupting their cellular integrity. This model makes use of a reporter mouse ubiquitously expressing the coral derived, photoconvertible, fluorescent protein Kaede (Tomura et al., 2008), which can be photoconverted from green to red after exposure to ultraviolet light with a wavelength of 350-400 nm. Kaede mice represent a powerful tool for tracking migratory intestinal TH17 cells to inflamed peripheral organs (Krebs et al., 2016; Magnuson et al., 2015; Morton et al., 2014). To track the fate of GALT cells in MCA205 tumor bearers, ileum, caecum and mesenteric LN were photoconverted at day 10 (D10) after subcutaneous tumor implantation and mice were killed 24 hours later for detection of photoconverted cells in various organs by flow cytometry (
To decipher the phenotype of the α4β7+ CD4+ T cells emigrating from the mesenteric LN, we performed a bulk-RNA sequencing of α4β7high CD4+ T cells compared with α4β7− CD4+ T cells purified from mLN of tumor bearers. The transcriptional profile of α4β7high CD4+ T highlighted—not only their Itga4 (encoding the α4β7 integrin) gut hallmark—but also a marked increase in gene products associated with Treg functions (such as Icos, Ctla4, Cd74, Mki67, P2rx7 (Daniel et al., 2010) and TH17 (Il17re, Tnfsf11, Il22) polarization, while the Tnfrsf9 T-cell costimulatory gene product (4-1BB) was significantly downregulated in this subset (
Next, we addressed how ATB-induced dysbiosis affected the exodus of enterotropic α4β7+ CD4+ T cells to the tumor draining LN using both imaging approaches combined to intracellular flow cytometry (
Neutralizing anti-MADCAM-1 antibodies also promoted the emigration of CD25high α4β7 CD4+ T (
Because photoconversion does not allow precise intracellular staining and high dimensional phenotyping, we dwelled into a deeper characterization of the gut emigrating cells performing Rhapsody-based single cell RNA sequencing of CFSE+ CD4+ T cells reaching the tdLN at 24 hrs post-CFSE injection of the mLN at day 4 after a 14 days ATB course (or no ATB) supplemented or not with E. clostridioformis (
Hence, we demonstrated the exit of enterotropic α4ρ7+ CD4+, most specifically Tr17 cells exhibiting immunosuppressive fingerprints, potentially harmful on the tumor immunosurveillance, from the GALT to tumor draining lymph nodes by two independent tracing methodologies, in conditions where the MAdCAM-1/α4β7 axis was compromised, either by anti-MAdCAM-1 mAb or by ATB-induced recolonization.
Given the reported immunosuppressive role of Tr17 cells during cancer immuno-surveillance (Blatner et al., 2012; Rizzo et al., 2018; Voigt et al., 2017) and the capacity of a healthy gut to retain these T cell subsets, we anticipated that genetic defects in MADCAM-1 addressin or 37 integrins or host neutralization of MADCAM-1 or α4β7 heterodimers by specific antibodies could interfere in the immuno-stimulatory capacity of therapeutic anti-PD-1 mAbs.
First, loss of ileal Madcam1 expression correlated with increased tumor size in MCA205 tumor bearing animals regardless of PD1 blockade (
Recolonization post-4 days ATB stop phenocopied madcam-1 gene deficiency inducing a 3-5 fold increase in the proportion of RORγt+ Treg (Tr17) in tumor beds (
The Tr17 recruitment promoted by PD1 inhibition was further increased when E. clostridioformis was supplemented by oral gavage after ATB stop, while in similar conditions, oral Lactobacillus reuteri failed to do so (
Given that IL-17 and IL-22 have direct and indirect pro-angiogenic and pro-tumorigenic effects (Lim and Savan, 2014; Voigt et al., 2017), neutralizing either one or both cytokines may circumvent the harmful effects of the recolonization post −ATB during PD1 inhibition. We undertook treating mice inoculated with a tumorigenic dose of WT or IL-22Rα1−/− deficient mammary 4T1 tumor cells by ip injections of anti-PD1 Abs in conjunction with anti-IL-17A neutralizing antibodies (
Therefore, we conclude that PD1 blockade facilitates the amplification and homing to tumors of TH17/Tr17 in the GALT in the setting of intestinal dysbiosis.
Re-establishing the fitness of the α4β7/MAdCAM-1 axis by external maneuvers may revert resistance to PD-1 blockade in tumor bearers treated with ATB or suffering from overt gut dysbiosis. Restoring gut eubiosis before starting PD1 blockade may represent a reasonable option. Short term ATB followed by fecal microbial transfer from a patient relapsing despite αPD-1 Abs caused ileal Madcam1 downregulation (
Hence, restoring intestinal eubiosis or neutralizing IL-17A represent potential options to counteract ATB-induced resistance to PD1 blockade.
Because intestinal tissue biopsies are difficult to obtain, friendly-using blood assays have been developed such as a sMAdCAM-1 ELISA to study the follow up of patients suffering from IBD and treated with anti-heterodimeric α4β7 antibodies (such as vedolizumab) (Holmer et al., 2020). However, soluble MAdCAM-1 levels have not also been monitored in mice. We found robust correlations between ileal madcam1 gene expression and serum soluble MAdCAM-1 (
Lastly, to directly show the relationship between soluble MAdCAM-1 serum levels and gut dysbiosis in advanced NSCLC patients, we performed a non supervised and supervised hierarchical clustering of the metagenomics species (MGS) defining the taxonomic composition of the intestinal microbiota according to the median of sMAdCAM-1 serum levels using shot gun metagenomics (MG) analyses in n=95 patients providing all the requested MG and serum pieces of information. First, there was a reduction in the MGS richness in MAdCAM-1low patients (segregated based on the median of all 112 NSCLC patients) (
Finally, selecting only the list of MGS found in mice (
Altogether, these findings indicate that soluble MAdCAM-1 is a surrogate marker of intestinal dysbiosis predicting resistance to PD1 blockade in cancer patients.
Our findings unveil the pivotal role of the ileum in hosts challenged by a tumorigenic process whereby the GALT keeps in check the emigration of gut tropic immunosuppressive TH22/TH17, among which up to 20% are Foxp3+CD25+ Treg cells. Both the gut and the TME are rich in cell attraction and homing molecules competing for circulating lymphocytes expressing their receptor or ligand pairs. Using tools enabling fate mapping from the GALT to tumor beds, we uncover that the fate decision for enterotropic T cells between gut residency and chemoattraction towards tumor deposits is—at least partly-controlled by the MAdCAM-1 addressin/α4β7 integrins molecular interactions.
First, ATB downregulated MAdCAM-1 expression in LP ileal venules and HEV of the mucosae and the mLN, favoring the ileal exodus of the α4β7+ Th17/Tr17 cells towards extraintestinal lesions. Secondly, MAdCAM deficiency or neutralization phenocopied the effects of ATB. Madcam-1 or #7 integrin gene deficient mice or animals receiving neutralizing Abs targeting these molecules failed to respond to ICI, by accumulating gut derived-α4β7+ Tr17/TH17 cells in the TME. Third, maneuvers aimed at restoring MAdCAM-1 expression on ileal HEV (such as A. muciniphila) or blocking IL-17A bioactivity compensated for the inhibitory effects of ATB in tumor immunosurveillance. Fourthly, enforced oral gavage with Enterocloster spp (such as E. clostridioformis) exacerbated the exodus of enterotropic Tr17 to tdLN. Finally, PD1 blockade tended to aggravate the expansion and/or priming of α4β7+ Tr17 cells in mLN and homing back to tumors, compromising cancer immunosurveillance. All these lines of evidence support the epidemiological studies pointing to a harmful effect of ATB prior to ICI administration, as opposed to during or after ICI start (Derosa et al. 2022). Consequently, we show that serum soluble MADCAM-1 is a biological proxy for ileal madcam1 gene expression and represents a reliable predictor of response to PD1 blockade, at least in NSCLC patients. Low sMADCAM-1 serum levels mirrored intestinal dysbiosis dominated by pro-TH17 bacteria (such as Veilonella parvula) or species from the Enterocloster genus associated with physiopathological disorders (Ghosh et al., 2020; Tsay et al., 2021; Zitvogel and Kroemer, 2021) (Derosa et al. 2022).
This demonstration follows prior evidence showing that gut derived-TH17 also control extraintestinal autoimmune (Magnuson et al., 2015; Morton et al., 2014; Wu et al., 2010) (Krebs et al., 2016; Lee et al., 2011b) or inflammatory/ischemic lesions (Benakis et al., 2016; Liesz et al., 2009). RORγt+ Treg harbour an exacerbated immunosuppressive phenotype compared with their lineage related-Treg with overexpression of gene products (such as Ctla4, Icos, Havcr2 (Sefik Science 2015) that we also found upregulated in our ATB-induced gut dysbiosis setting. Moreover, we identified gut-specific Treg features in our Tr17 fingerprint previously described in landmark papers (Miragaia et al., 2019; Munoz-Rojas and Mathis, 2021), TH17 related blueprints (Yosef et al., 2013) as well as immunosuppressive traits functionally relevant in tumor immunosurveillance (such as Dusp2/PCA1 (Dan Lu et al., 2020), PIK3ip1 (Uche et al., 2018).
This exemplification in cancer bearers is in line with previous studies performed in patients or macaques suffering from inflammatory bowel disease, reporting an increased recirculation of extraintestinal α4β7+ Treg and central memory T cells or CCR6+ CD4+ T lymphocytes during treatment with the anti-α4β7 antibody vedolizumab (Calenda et al., 2018; D'Haens et al., 2018; Fischer et al., 2016). Our findings have important clinical implications. So far, vedolizumab was considered as a safer alternative to TNFα inhibition in cancer patients suffering from IC-induced autoimmune colitis because of their gut restricted-mode of action (Sandborn et al., 2016). However, prospective studies should monitor microbiome composition, gut or soluble serum MAdCAM-1 expression, as well as the recirculation of CCR9 α4β7+Treg, Tr17 and Th22 cell subsets to correlate these novel parameters with the clinical outcome and toxicity profiles in patients treated with ICI. Finally, the field of fecal microbial transplantation may evolve towards guiding selection of donor feces based on their capacity to normalize MADCAM-1 levels in the cancer bearing recipients.
These findings fuel the concept that cancer bearing patients can exacerbate their cancer-associated ileopathy (Yonekura et al 2022) by taking antibiotics that alter one of the most potent gut immune checkpoint MADCAM-1. We surmise that ATB-induced dysbiosis and most specifically recolonization post-ATB course is a major regulator of gut immune checkpoint MADCAM-1. Several molecular cues may link gut dysbiosis with the alteration of GALT MADCAM-1 such as perturbations of the bacteria-induced metabolism of biliary salts (Campbell et al., 2020; Song et al., 2020), and activation of the sympathetic nervous system for instance (Schiller et al., 2021; Yan et al., 2021), that deserve further investigations.
In Example 1, we showed, in two independent cohorts of advanced non small cell lung cancer patients (NSCLC), that serum soluble MAdCAM-1 level is a marker of resistance or sensitivity to an immuno-oncology (I-O) therapy, wherein a reduced level of serum soluble MAdCAM-1 is a marker of resistance to the I-O therapy.
Here, we extend this finding to another cancer type: the kidney cancer treated with nivolumab in second line therapy (after failure of tyrosine kinase inhibitors) (Table 8).
Overall, 212 patients from the NIVOREIN cohorts were enrolled to receive anti-PD1 mAb (Nivolumab, EMS). 31 were antibiotics users (between day −60 and day +42 after treatment initiation) and 176 did not take antibiotics. The ELISA monitoring of serum soluble MAdCAM-1 at baseline was performed, revealing a median (min,max) of 88.8 ng/ml (19.1, 174.7) (and a mean±SD of 88.4±27.7 ng/ml) for these 212 metastatic clear cell renal cell carcinoma (RCC) patients. As shown for NSCLC patients, antibiotic uptake (ATB) tended to decrease the concentrations of sMAdCAM-1 (from a median of 88.7 ng/ml in those patients who refrained from taking ATB down to 76.3 ng/ml in those who took ATB). Of note, this is underestimated since the vast majority of these ATB+patients (23/31) took ATB after initiation of anti-PD1 Ab (which is less clinically significant).
The overall survival (OS) was analyzed according to the median value of sMAdCAM-1 baseline levels right before anti-PD1 Ab initiation, for the whole cohort of 212 patients. The Kaplan Meier OS curves of RCC patients show that those presenting a sMAdCAM-1 serum level <88.8 ng/ml died in 72/106 cases, while the mortality rate was 36/106 in those who had a sMAdCAM-1 serum concentration 88.8 ng/ml (p<10e-4) (
The progression-free survival (PFS) was analyzed according to the median value of sMAdCAM-1 baseline levels right before anti-PD4 Ab initiation, for the whole cohort of 212 patients. The Kaplan Meier PFS curves of RCC show that patients presenting a sMAdCAM-1 serum level <88.8 ng/ml progressed in 99/106 cases, while progression rate was 86/106 in those who had a sMAdCAM-1 serum concentration ≥88.8 ng/ml (p<10e-4) (
As for objective response rates (ORR), NIVOREN trial (N=729 RCC patients) recorded an objective response rate of 20.8% (1.2% of complete responses and 19.6% of partial responses, 79.2% of progressive diseases (PD)). Interestingly, the subgroup of 212 RCC patients for whom sMAdCAM-1 serum levels were available, was subdivided according to ATB uptake Yes/no and to sMAdCAM-1 (≥88.8 versus <88.8 ng/ml). First, subgroup analysis of best objective response according to sMAdCAM-1 revealed that the percentages of PD was 36.7% vs 62.6% in patients with sMAdCAM-1≥88.8 versus <88.8 ng/ml respectively (p=0.0003) (Table 10). Next, when we considered only patients who did not take ATB, the same subgroup analysis of best objective response according to sMAdCAM-1 revealed that the percentages of PD was 38.6% vs 60.5% in patients with sMAdCAM-1≥88.8 versus <88.8 ng/ml respectively (p=0.0044). Thirdly, when we considered RCC patients who took ATB, this subgroup analysis of best response rate according to sMAdCAM-1 revealed that the percentages of PD was 36.4% vs 76.9% in patients with sMAdCAM-1≥88.8 versus <88.8 ng/ml respectively (p=0.0446). In the ATB-treated subgroup, the percentage of complete responders was 18.2% versus 7.7% in patients with sMAdCAM-1≥88.8 versus <88.8 ng/ml respectively (Table 11 below).
Altogether, these findings demonstrate that high levels of serum soluble MAdCAM-1 are associated with increased response to PD1 blockade in second line immunotherapy of RCC, and more so in patients who took antibiotics. The threshold of 88 ng/ml is valid for second line but is not representative of first line patients. For routine usage, the median of a cohort of 1 L or 2 L therapy must be considered.
Finally, we sought to triplicate the predictive value of sMAdCAM-1 in a subcohort of a Phase 3a trial enrolling pre-treated metastatic bladder cancer patients to receive a fixed dose of 1500 mg of Durvalumab iv (anti-PD-L1 Abs)/4 weeks. In this STRONG study, 79 patients only were analyzed, with an overall median OS of 5.79 months, and a median PFS of 2.79 months. We selected the extreme of the STRONG cohort, 30 Elite patients (OS>6 months, PFS>5 months, PR+CR only) and rapid progressors (n=49, with OS<6 months and PFS<5 months and PD at the first CT scan). In this cohort, the MAdCAM-1 median (min,max) was 158.8 ng/ml (53.41, 244.36). In
Altogether, these findings demonstrate that high levels of serum soluble MAdCAM-1 are associated with increased response to PD1/PDL-1 blockade in second line immunotherapy of RCC, in second line metastatic bladder tumors and in 1 L and 2 L NSCLC treated with pembrolizumab regardless of ATB uptake.
α4β7: alpha4beta7, Abs: antibodies, ACS: Ampicillin, Colistin, Streptomycin, ATB: antibiotics, ATRA: all-trans retinoic acid, CCL: Chemokine ligand, CCR: Chemokine receptor, CD: Cluster of differentiation, cDC1: classical dendritic cell type 1, CFSE: Carboxyfluorescein succinimidyl ester, cLN: contralateral lymph node, CSF-1: colony stimulating factor 1, CTL: cytotoxic T lymphocyte, FMT: fecal mucosal transfer, FoxP3: forkhead box P3, GALT: Gut-associated lymphoid tissue, HEV: high endothelial venule, IBD: inflammatory bowel disease, ICI: Immune checkpoint inhibitors, IFNγ: Interferon γ, LP: lamina propria, mABs: monoclonal antibodies, MAdCAM-1: Mucosal Addressin Cell Adhesion Molecule-1, mLN: mesenteric lymph node, PP: Peyer's patch, RA: retinoic acid, RORC: RAR-related orphan receptor C, SLO: secondary lymphoid organ, tdLN: tumor draining lymph node, TH17: T-helper 17-CD4-T-cell, TME: tumor microenvironment, TNFα: Tumor necrosis factor alpha, Treg: regulatory T-cell, Tr17: RORγt+ Treg-CD4-T-cell.
| Number | Date | Country | Kind |
|---|---|---|---|
| 21305846.4 | Jun 2021 | EP | regional |
| 21306817.4 | Dec 2021 | EP | regional |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/EP2022/066929 | 6/21/2022 | WO |
