Methods and compositions for treating cancer

Abstract

The present invention consists in combining low dose intestinal radiation (ILDR) with immunotherapy or other antineoplastic treatments, in the ideal context of intestinal presence or prevalence of Christensenellaceae family members, especially for solid cancers at an advanced stage (refractory to first line chemo or immunotherapy). The invention also pertains to a bacterial composition,e.g. comprising Chistensenella bacteria, for use in the treatment of cancer in a patient, in combination with intestinal low dose radiotherapy (ILDR) and an antineoplastic treatment.

Description

[0001] Methods and compositions for treating cancer

[0002] FIELD OF THE INVENTION

[0003] The present invention relates to the field of cancer immunotherapy.

[0004] BACKGROUND OF THE INVENTION

[0005] Metastatic disease is the main cause of cancer-associated death. The clinical management of advanced cancers requires multiple treatment modalities, including surgery, radiotherapy (RT), chemotherapy, precision medicine and immunotherapy. Immunotherapy based on immune checkpoint blockade (ICB) has shown promising results in metastatic cancer and became an integrated component of their multidisciplinary care (Ribas et al, 2018). The therapeutic efficacy of ICB depends on miscellaneous cell-autonomous and cell-extrinsic factors (Pitt et al, 2016) . RT can also exhibit immunostimulatory properties, triggering immunogenic tumor cell death. Animal studies suggested that combination of RT, including stereotactic ablative radiotherapy (SABR), and ICB could mediate synergistic anti-tumor immune responses, eventually leading to “abscopal” effects described in some cases of local irradiation (Chen et al., 2025). Abscopal activity refers to the capacity of local irradiation to induce regression of unirradiated distal tumor lesions. Hundreds of trials have been launched to optimize combinational regimens of RT+ICB in a plethora of malignant diseases. Studies identified the interferon (IFN) signaling cascade as a biomarker of efficacy of the combinatorial regimen (Minn and Wherry, 2016). Later, additional clinical studies supported the rationale of this combinatorial approach, while other trials were unsuccessful, specifically in patients affected by head and neck, advanced cutaneous tumors and ICB-refractory non-small cell lung cancer (NSCLC). These conflicting results suggest the need for a better understanding of the mechanisms underlying the coordinated effects of both treatment modalities.

[0006] RT may induce intestinal off-target effects on the intestinal epithelial barrier. Off-target irradiation of the gut induces mucosal damage and dysbiosis, eventually causing enteritis (Hauer- Jensen et al, 2014). RT-induced enteritis influences gut microbiota composition, which reciprocally impacts its severity (Guo et al 2020). Severe enteritis is detrimental to patient quality of life, therapeutic tolerability, and therapeutic benefit. Hence, RT of the abdominopelvic region is subjected to specific dose constraints (Jadon et al 2019). However, the gut microbiota has emerged as a key modulator of immunotherapy, including ICB, CAR-T cells, hematopoietic stem cell transplantation and RT. The metagenomics-based taxonomic composition at diagnosis influences clinical outcome to ICB across geographical sites and cancer types (Routy et al, 2018; Vetizou et al 2015; Gopalakrishnan et al, 2018; Zitvogel et al, 2018). Intestinal commensals shape systemic immunity and the tumor microenvironment at distal sites through various mechanisms (Pitt et al, 2016; Morad et al, 2022; Derosa et al, 2021 ; Fluckiger et al, 2020; Fidelle et al, 2023). Besides their indirect activity on the immune system, distinct bacteria species display cell autonomous effects on irradiated tumor cells. Hence, Bacteroides species attenuate tumor response to chemoradiotherapy in rectal cancers by providing nucleotides accelerating DNA damage repair (DDR) (Teng et al, 2023), while Lactobacillus iners, a L-lactate producing species, confers chemotherapy and radiation resistance in cervical cancer cells through metabolic rewiring (Colbert et al2023).

[0007] SUMMARY OF THE INVENTION

[0008] To decipher the role of the host-microbe interaction in the toxicity and efficacy of the SABR+ICB combinatorial regimen, the inventors investigated gut-specific off-target effects of irradiation, called “intestinal low-dose irradiation” (ILDR) henceforth, on the metagenome, metabolome and inflammatory tonus, as well as on the clinical outcome of metastatic cancer patients. They found that ILDR was a strong clinical hallmark of long-term benefit to SABR+anti-PD-L1 Abs. This finding was validated in preclinical models where ILDR was indispensable for the abscopal effects of ICB on unirradiated tumors (Chen et al, 2025). This involved microbial, metabolic and immune effects culminating in the recruitment of non exhausted effector CD8+T cells into the tumor bed. Christensenella minuta markedly increased the efficacy of ILDR by mobilizing dendritic cells (DC) from the mesenteric to tumor draining- lymph nodes and correlated with plasma levels of secondary bile acids increasing antigen cross-presentation by DC.

[0009] The present invention consists in combining low dose intestinal radiation (ILDR) with immunotherapy or other antineoplastic treatments, in the ideal context of intestinal presence or prevalence of Christensenellaceae family members, especially for solid cancers at an advanced stage (refractory to first line chemo or immunotherapy).

[0010] The invention also pertains to a bacterial composition, e.g. comprising Chistensenella bacteria, for use in the treatment of cancer in a patient, in combination with intestinal low dose radiotherapy (ILDR) and an antineoplastic treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1. Stool metagenomics and serum metabolomics profiling reveal beneficial commensals and their metabolites in responders to SABR-aPD- L1.

[0012] (A) Differential abundance analysis performed via a Mann-Whitney test for the fecal microbiome of Elite (n=6) versus non-Elite (n=18) patients at V1. Benjamini- Hochberg procedure was applied to control for FDR. None of the tested species-level genome bins (SGBs) showed a Q < 0.2. We report here the associations with a noncorrected P < 0.05. The bars show the standardized mean difference (SMD). (B) Comparisons of relative abundances of four key Metagenomic Species (MGSs) in responders (Rs) (upper panel: OS12, lower panel: Elite), non-responders (NRs) and healthy volunteers (HVs, n = 5234).

[0013] Figure 2. Effects of low-dose intestinal irradiation on the antitumor efficacy of PD-L1 blockade in tumor-bearing mice and the key role of C. minuta in the synergy.

[0014] (A) Experimental setting for the combination of ILDR and anti-PD-L1 Abs in MC38 colon cancer-bearing mice and timelines of sample harvesting (left panel). MC38 bearing C57BL / 6J mice were subjected to intraperitoneal (ip) injections of isotype control Ab versus aPD-L1 Ab and abdominal ILDR (with increasing dosing of 0.25, 1 or 4Gy). The image on the right illustrates a representative murine tomography showing 1Gy ILDR by three 220 kV X-ray beams with isodose curves on the transverse plane. The tumor (indicated in dashed white circle) was spared from all beams. (B). Tumor growth kinetics for each individual mouse in the control and aPD-L1 Ab groups are displayed in the left panel, while the individual growth curves for ILDR and ILDR + aPD- L1 groups are depicted in the middle panel. Kaplan-Meier survival curves and Log-rank test of all murine subgroups and numbers of complete tumor eradications are shown in the right panel. Each experiment comprised 5-6 mice / group and was repeated 3 times, the graph gathering all three experiments. (C) Kaplan-Meier survival curves and log-rank test of mice subjected to aPD-L1 Ab in combination with a range dose of ILDR (0-4Gy). N = 5-15 from at least two independent experiments. (D) Kaplan-Meier survival curves and log-rank test of mice subjected to aPD-L1 Ab in combination with either 1Gy ILDR or liver RT. N = 9 from least two independent experiments. (E-F) Mice were subjected to FMT with fecal samples from 8 patients, respectively prior to aPD-L1 with or without ILDR. Mice were grouped according to their benefit from the additional ILDR: with benefit (N = 5, E) vs without (N = 3, F). A representative growth curve is shown on the left with the Kaplan-Meier survival of curves of all mice on the right. (G) Mice subjected to oral gavage of NaCI or C. minuta 601 (on day-4, 7, 9 and 11) were treated with aPD-L1 + ILDR. Tumor growth kinetic and survival curves are shown. (H) Tumor growth kinetics of mice colonized with another strain of C. minuta (spp-2), treated with the combinational treatment. (I) Tumor growth kinetics of mice subjected to oral gavage of a variety of bacterial species, including O. splanchnicus (two different strains), B. fragilis, B. thetaiotaomicron, P. distasonis (two different strains, also referred to Figure 9),. ns: P > 0.05; * P < 0.05; ** P < 0.01 ; *** P < 0.001 ; **** P < 0.0001. See also Figures 7-9.

[0015] Figure 3. mregDC-CD8+T cell-dependent effects of the combinatorial regimen ILDR + aPD-L1.

[0016] (A) Anti-Isotype Ctrl, anti-CD4- or anti-CD8-depleting Ab were injected ip into MC38-bearing mice undergoing aPD-L1 AB + ILDR 1Gy on day 6, 8 10 and 15 according to the same experimental setting detailed in Figure 2A. Tumor sizes at sacrifice are depicted. (B-E) MC38-bearing mice were subjected to isotype Ab or aPD- L1 with 0-4 Gy ILDR. Tumors were harvested at day 11 and analyzed by flow cytometry for immune cell profiling. The percentages of CD3+CD8+ T cells within total live cells are shown in (B). (C) The percentages of TCFT TOX+cells and TCFT PD1+cells among CD44+CD8+T (also refer to Figure 9F for a representative dot plot) among CD44+CD8+ T cells. (D) The percentages of a4p7+Tregs (CD3+CD4+CD25+Foxp3+) within CD45+cells and their ratio with enterotropic a4p7+CD8+T cells are depicted (E). N = 9-14 mice / group from two independent experiments. (F) The percentage of mregDC (gated as lineage negative [CD3-CD19-NK1.TLy6G’Siglect F’] CD11c+MHC ll+PD-L1+CD40+) in CD45+cells in TdLNs or mLN collected from mice at 3 days after aPD-L1 with or without ILDR. (G) At 2 days after aPD-L1 ± ILDR, mice were subjected intra-mLN injection of carboxyfluorescein succinimidyl ester (CFSE) (100mM). Twenty-four hours after surgery, mice were sacrificed and the percentage of PD-L1+DCs in CFSE- or CFSE+ CD45+cells were assessed (model on the right, results on the left). (H) Mice were treated with aPD-L1 ± ILDR ± oral gavage of C. minuta 601. The infiltration of CCR7+or PD-L1+CD40+CCR7+DC (gated on CD45+cells or DC) in TdLN was evaluated and compared. (I) Volcano plot aligning the significant differences between aPD-L1 vs aPD-L1+ ILDR 1Gy subgroups of mice in the relative abundances of plasma metabolites at day 11 in mice (P < 0.05). N = 10 mice / group in 2 independent experiments pooled together. (J-L). High content in vitro screening of metabolites onto a DC / T cell assay. The inducible immortalized dendritic cell line (i ni DC) was differentiated into de-iniDCs upon removal of dexamethasone and doxycycline and subjected to a dose range (0-6Gy) of IR (intervention 1), or a metabolite library of 96 compounds including bile acids / carnitines / sterols (at two concentrations i.e 2.5pg and 5pg / ml) for 16 hours (intervention 2), then pulsed with OVA protein (1mg / mL) for 4 hours and incubated with CTL (B3Z hybridoma cells) for 16 hours. ELISA determination of mlL-2 concentrations is depicted. The scheme is displayed in (J) and the results are shown in (K-L). IL-2 concentrations in intervention 2 were normalized to the mean values of dimethyl sulfoxide-treated controls. Data represent means ± SD. Comparisons of means between 3 groups were performed using the Kruskal-Wallis H test, ns: P > 0.05; * P < 0.05; ** P < 0.01 ; *** P < 0.001 ; **** P < 0.0001. See also Figures 9-11.

[0017] Figure 4. C. m / nuta-associated metabolites impact DC crosspresentation function.

[0018] (A-B) The lactic acid and relevant metabolites in mice subjected to isotype control Ab or aPD-L1 ± 1 or 4Gy ILDR (plasma samples collected from mice in Figure 6B). Heatmap in (A) and dot plots with comparison in (B). (C-D) (C) The prevalence (right) and relative abundance (left) of Christensenella species in fecal samples collected from mice in Figure 2B. (D) Volcano plot aligning the significant differences between subgroups of mice with detectable C. massiliensis versus without in the relative abundances of biliary salts at day 11 in mice (Mann- Whitney U test, P < 0.05 without multiple comparison correction). (E) The impact of treatment with a range concentration (0-50mg / ml) of Deoxycholic acid (DCA) or Ursodeoxycholic acid (UDCA) on the cross-presentation capacity of DCs (refer to Figure 3J). IL-2 concentrations in experimental groups were normalized to the mean values of the control group (0 mg / mL) for fold changes, n = 3-8 from two independent experiments (F-G) (F) The prevalence (right) and relative abundance (left) of C. minuta in fecal samples collected from SABR patients at multiple timepoints (refer to Figure 1A). (G) Volcano plot aligning the significant differences between subgroups of fecal samples with detectable C. minuta versus without in the relative abundances of biliary salts in serums contemporarily from SABR patients (Mann- Whitney U test, P < 0.05 without multiple comparison correction).

[0019] Figure 5. 1Gy ILDR restored therapeutic sensitivity in ICB refractory metastatic cancer patients enrolled in a perspective phase II trial.

[0020] (A) The design of the phase II ILDR perspective trial (NCT06076135). (B) A representative dosimetric graph of 1Gy ILDR. (C-E) (C) The waterfall graph of the best tumor response (by RECIST 1.0) of the first 10 patients consecutively enrolled in this study (also refer to Table 4). (D) The spider graph of tumor response overtime after treatment. (E) The swimlane graph of patient response to treatment. (F) Representative CT images of a scalp metastasis before and after ILDR + 4 cycles of PD-1 blockade. (G) Representative CT images of a mediastinal lymph node metastasis taken before and after ILDR + ICB.

[0021] Figure 6. Metagenomics analysis of fecal samples and metabolomic analysis of serum samples in OS12 R and NR patients, related to Figure 1(A).

[0022] Differential abundance analysis performed via a Mann-Whitney test for the fecal microbiome of OS12 R (n=14) versus OS12 NR (n=10) patients at V1. Benjamini-Hochberg procedure was applied to control for FDR. None of the tested SGBs presented Q < 0.2. We report here the associations with a non-corrected P < 0.05. The bars show the standardized mean difference (SMD).

[0023] Figure 7. Abdominal tumors displayed greater therapeutic benefit from SABR compared with cervical lesions in an autochthonous mammary tumor mouse model and metagenomic analysis of donor fecal samples used in the FMT experiment, related to Figure 2.

[0024] Female, 6-7 weeks old C57BL / 6J mice were subjected to subcutaneous implantation of slow-release (90 days) of medroxyprogesterone acetate pellets (50mg) and oral gavage of 7,12-dimethylbenz[a]anthracene (1 mg per week for 7 weeks) for induction of autochthonous mammary tumors. Mice with primary tumor area reaching 15-45 mm2were randomized into subgroups to receive mock or stereotactic ablative radiotherapy (SABR, scheme in A, created by Biorender): 20Gy x 2F or 20Gy x 3F. Following treatment, mice were monitored for tumor growth and the presence of secondary tumors. Mice were euthanatized when reaching ethical endpoints (e.g. the cumulative tumor surface > 180-200 mm2, or in the presence of evident signs of toxicity or stress). Mice were retrospectively categorized into two groups according to the location of primary (abdominal vs cervical) tumors to which SABR was delivered. The overall survival (OS) and the time to secondary tumor (TTS, a surrogate marker of the abscopal effect) was monitored in these two groups of mice (mock in B, SABR in C) and represented in Kaplan-Meier curves analyzed using the log-rank method (B-C, OS in the left panels, TTS in the right panels). The results are pooled from at least two independent experiments and the number of mice in each subgroup is indicated

[0025] Figure 8. Kinetics of the taxonomic fecal microbiota composition in mice according to ILDR scheduling, related to Figure 2.

[0026] (A) Longitudinal comparison of the relative abundance of 3 key species in murine stools as assessed by qPCR comparing two different time points for each line. Results are shown as Iog2 transformed ratios of species abundances. n=5-19 from at least two independent experiments. (B) Relative abundance of Ruminococcus bromii over time in patients with (right, n=6) or without (left, n=18) ILDR. Statistical significances between V3 and V4 are indicated. The data represent mean±SEM. (C) The expression of Reg3g or Defa5 (normalized to Ppia) in ileal tissues collected at 3 days after ILDR + aPDL1 Ab vs aPD-L1 Ab alone. n=7-8 / group from two independent experiments. (D) The expression of Reg3g or Defa5 in ileum-derived organoids (with 100pM noradrenaline in the culture medium) at 24 hours after 0-4Gy IR. n=5 / group from two independent experiments. Data are shown as mean±SD in I and J. Statistical analysis: Wilcoxon matched-pairs signed rank test (A). Kruskal-Wallis H test (B, C and D). Mann-Whitney II test (D). P values <0.05 were considered significant. * P<0.05, ** P<0.01.

[0027] Figure 9. P. distasonis compromised the synergistic antitumor effect of ILDR + aPD-L1 combinatorial regimen, related to Figure 2.

[0028] (A) Kaplan-Meier survival curves of patients segregated into two groups according to the relative abundance of fecal P. distasonis SGB1934 at V1. Cox regression and Log rank test. The number (#) of patients at risk for each group is indicated. (B) Spearman correlation coefficients between the relative abundance of Parabacteroides distasonis SGB1934 and the top 2 pathways involved in nucleotide metabolism: PWY-7228 guaosine nucleotides biosynthesis I and PWYO-1296 pathway purine ribonucleotides degradation (FDR correction via BH procedure). (C) Mice were subjected to oral gavage with P. distasonis or PBS at day 5 and 7 (as in Figure 2A), in addition to ILDR 1Gy + aPD-L1. Relative abundance of P. distasonis in orally fed mice, as assessed by specific qPCR of fecal samples collected at day 7 post gavage is shown in the left. Kaplan-Meier survival curves and log-rank test of mice and numbers of complete tumor eradication are shown in the right. A typical experiment comprising 5 mice / group, was performed 2-3 times and the results of pooled experiments is depicted. (D) The relative abundance of deoxyguanosine in plasma metabolomics from all mice subgroups is shown, each dot representing one mouse. A representative experiment comprising 5 mice / group is depicted out of 2 yielding similar results. (E) The percentage of a4p7+ Tregs in CD45+ in tumors collected on day 11 from mice treated with aPD-L1 + I DLR with oral gavage of NaCI or P. distasonis. The comparisons of means of absolute values were performed using the Mann-Whitney U test between 2 groups and the Kruskal-Wallis in > 3 groups. * P < 0.05, ** P < 0.01 , *** P < 0.001 , **** p < 0.0001. (F) Crypt stem cell-derived murine enteroids were subjected to a dose (0-6Gy) range of irradiation (IR). Fortyeight hours later, the culture supernatant was harvested and the concentration of murine IL-7 was assessed by ELISA. n=17 for each dose from 3 independent experiments. (G) MC38-bearing C57BL / 6J mice were subjected to therapy with i.p. isotype control Ab, aPD-L1 Ab, aPD-L1Ab plus 1Gy ILDR or aPD-L1 Ab plus 4Gy ILDR. Mice were sacrificed at day 11 and their ileal tissues were collected and analyzed by RT-qPCR for II7 expression (normalized to Ppia), N=10-15 mice / group from two independent experiments. (H). IL-7R neutralization with intraperitoneally injected specific antibodies given on day 5, 7, 9 and 13 (referto Figure 2A for scheme). The Kaplan-Meier survival curves of MC38-bearing mice subjected to aPD-L1 plus ILDR with isotype control or alL-7R Ab are depicted for two combined experiments (N=14-15 mice / group). Data are shown as mean±SD in C-G. Statistical analysis: Log-rank test (A, C right and G). Spearman’s correlation (B). Mann- Whitney U test (C left and D). Kruskal-Wallis H test (F). One-way ANOVA (G). P values <0.05 were considered significant. * P<0.05, ** P<0.01 , *** P<0.001.

[0029] Figure 10. Intratumoral infiltration of immune cell subsets in mice subjected to various therapies, related to Figure 2.

[0030] (A-D) MC38-bearing C57BL / 6J mice were subjected to isotype control Ab, aPD-L1 Ab, aPD-L1 Ab + 1Gy ILDR or aPD-L1 Ab + 4Gy ILDR. Mice were sacrificed and their tumors were analyzed by flow cytometry for evaluation of immune cell subsets. (A) The percentage of Tregs in CD4+T cells. (B)The percentage of PD1+TCF1+(bottom) or PD1+(up) cells in CD44+CD8+T cells. (C) The percentage of CXCR6+cells in TOFT PD1+CD44+CD8+T cells. (D) The percentage of a4p7+immune cell subsets in CD45+cells, including CD8+T cells, CD4+T cells and Tr17 cells. N = 10-15. Each dot represents one tumor and mouse. Data represent means ± SD. Two experiments were pooled and shown. Comparisons of means between subgroups groups were performed using the Kruskal-Wallis H test. * P < 0.05; ** P < 0.01 ; **** P < 0.0001

[0031] Figure 11. Phenotype and quantification of dendritic cells in mesenteric lymph nodes and tumor-draining LNs in mice subjected to various therapies, related to Figure 2.

[0032] (A-B) The percentage of PD-L1+ or CD40+ DCs in total DCs (top) or CD45+ cells (bottom) in tumor-draining lymph nodes (TdLNs) (left panels) or mesenteric LNs (mLNs) (right panels) is shown (A) and (B), respectively. (C -D) The proportion of PD-L1+CD40+ DC in total DC in TdLNs (C) or mLNs (D) is displayed, respectively. The results of a typical experiment containing 7-8 mice / group are depicted as mean±SD in B-D. Statistical analysis: Mann-Whitney U test (B-D). P values <0.05 were considered significant, ns P>0.05, * P<0.05, ** P<0.01. Tables

[0033] Table 1. The mean prevalence and relative abundance of significantly different species between Elite and non-Elite patients at V1.

[0034] Table 2. The mean prevalence and relative abundance of contrasting species between OS12 R and NR patients at V1.

[0035] Table 3. ANOSIM and PERMANOVA metrics and related P values from PCoA analysis of p diversity of microbiota in murine fecal samples collected at V2, V4 and V5.

[0036] Table 4. Clinical characteristics of the first 10 patients enrolled in the ILDR perspective trial

[0037] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0038] In one aspect, the invention relates to a composition comprising live bacteria belonging to the Christensenellaceae family, for use in the treatment of cancer in a patient, in combination with intestinal low dose radiotherapy (I DLR) and an antineoplastic treatment.

[0039] The composition according to the invention enhances the efficacy of I DLR and the antineoplastic treatment and improves tumor control, especially in patients with a refractory cancer.

[0040] Christensenellaceae is a recently described family within the Firmicutes phylum of Bacteria. It comprises the genus Christensenella. Thus, in one aspect, the composition according to the invention comprises live bacteria of the genus Christensenella.

[0041] The genus of Christensenella comprises the species Christensenella minuta, Christensenella massiliensis, Christensenella timonensis, Christensenella hongkongensis and Christensenella intestinihominis. Furthermore, Christensenellaceae bacterium SGB14127 also belongs to the genus of Christensenella. Thus, in a preferred embodiment, the composition according to the invention comprises live bacteria of at least one of the species Christensenella minuta, Christensenella massiliensis, Christensenella timonensis, Christensenella hongkongensis, Christensenella intestinihominis and Christensenellaceae bacterium SGB14127. In some embodiments, the composition comprises live bacteria from at least two, at least three or at least four different Christensenella species. In a particularly preferred embodiment, the composition according to the invention comprises live bacteria of the species Christensenella minuta, such as Christensenella minuta 601 deposited at the Collection Nationale de Cultures de Microorganismes (CNCM, Institut Pasteur, Paris) under the reference 1-6155.

[0042] In some embodiments, the composition further comprises, in addition to live bacteria from the Christensenellaceae family, an archea, preferably Methanobrevibacter smithii and / or Ruminococcus bromii bacteria.

[0043] Prerably, the composition according to the invention is a live biotherapeutic product (LBP) comprising at least one cultured bacterial strain. As used herein, the term “live biotherapeutic prodcut (LBP)” is defined as a a biological product that comprises live microorganisms, is applicable to the prevention, treatment or cure of a disease or condition in a human being and is not a vaccine or gene therapy agent.

[0044] In one embodiment, the composition according to the invention comprises a consortium of bacteria (synthetic microbiome consortium). The term “consortium of bacteria” as used herein refers to a mix of several bacterial species obtained from in vitro cultures.

[0045] According to the invention, the consortium of bacteria may comprise live bacteria of the Christensenella genus such as described above, live bacteria of Akkermansia genus such as Akkermansia massiliensis and Akkermansia muciniphila, and live bacteria selected amongst Ruminococcus bromii, Faecalibacterium prausnitzii, Ruminococcaeae, Prevotella copri, Alistipes shahii, Alistipes finegoldi, Eubacterium ventriosum and other bacterial species known to be beneficial for a patient’s response to immunotherapy, such as SIG 2 bacteria listed in WO2024 / 094817.

[0046] In one embodiment, the composition according to the invention comprises fecal material (including microbes) and is formulated to perform fecal microbiota transplant (FMT). The fecal microbial composition is preferably obtained (directly or indirectly) from a stool sample from (a) healthy individual(s) or (a) responder(s) to a treatment with an anti-neoplastic treatment such as immuno-oncology therapy. The fact that the fecal microbial composition can be obtained indirectly from a healthy individual’s stool sample means that banks of fecal microbial material may be created, with possible mixes of stool samples, and possible creation of “standard healthy fecal microbial compositions”, possibly adapted to certain characteristics of patients (age, ethnic origin, food regimen etc.). Such compositions for fecal microbiota transplantation (FMT) can also be enriched with other beneficial bacterial species, as described above. Several ways of conditioning fecal microbial material and conducting FMT have been described and are currently developed, and the skilled artisan is free to choose appropriate techniques for preparing the fecal microbial composition according to the invention, which can be freshly-prepared liquid, freeze-dried material or any other conditioning, e.g., by addition of suitable buffers such as saline solution or glyercol.

[0047] In one embodiment, the bacterial composition according to the invention is supplemented with metabolites. In a preferred embodiment, the metabolites are selected from ursodeoxycholic acid (LICDA), secondary bile acids (e.g., deoxycholic acid and taurolithocholic acid), arginine, tryptophane, indole proprionate, indole acetate, indole carboxylate, indole-3-carbinol and mixtures thereof.

[0048] As used herein, the term “antineoplastic treatment” relates to any treatment given for the treatment of cancer, in particular immuno-oncology therapy and chemotherapy.

[0049] In preferred embodiments, the antineoplastic treatment to be used together with the composition according to the invention comprises immuno-oncology (I- O) therapy. Immuno-oncology (l-O) therapy is herein defined as any immunotherapy applied for the treatment of cancer. Types of l-O therapy include, but are not limited to, immune checkpoint blockade (ICB), T-cell transfer therapy, monoclonal antibody therapy, immune system modulator therapy and cancer vaccines. In a particularly preferred embodiment, the l-O therapy comprises immune checkpoint blockade (ICB), such as anti-PD1 / PD-L1 / PD-L2 Ab-based therapy and / or anti-CTLA4 Ab-based therapy. In a further particularly preferred embodiment, the ICB comprises or consists of anti-PD- 1 antibodies (Ab) (such as nivolumab, pembrolizumab, cemiplimab, dostarlimab, retifanlimab, toripalimab, tislelizumab, sintilimab, serplulimab, camrelizumab and penpulimab), anti-PD-L1 Ab (such as atezolizumab, durvalumab, avelumab, sugemalimab and cosibelimab), anti-PD-L2 Ab and / or anti-CTLA4 Ab such as ipilimumab. In particular, it can be an anti-PD1 and / or anti-CTLA4 monoclonal antibody.

[0050] An “anti-PD1 / PD-L1 / PD-L2 Ab-based therapy” herein designates any drug that antagonizes PD1 or PD-L1 or PD-L2. Although the currently used drugs antagonizing PD1 or PD-L1 or PD-L2 are monoclonal antibodies, other molecules specifically binding to PD1 , PD-L1 or PD-L2 could be used for the development of future ICB such as, for example, antibody fragments or specifically designed aptamers. Of course, the phrase “anti-PD1 / PD-L1 / PD-L2 Ab-based therapy” encompasses any therapy with active molecules that antagonize PD1 or PD-L1 or PD-L2. The same applies to anti-CTLA4 Ab-based therapy.

[0051] In other embodiments, the antineoplastic treatment comprises chemotherapy such as oxaliplatinum based-cytotoxicants, immunogenic oxaliplatinum based therapy, anthracyclins, FOLFIRINOX (5 Fluoro-uracile Irinotecan Oxaliplatine, 5 Fluoro-uracile Campto Eloxatine), taxanes and any protocol, chemotherapeutic agent or combination thereof for the treatment of colon, pancreas, cervix, anal or pelvic carcinoma.

[0052] In some embodiments, the antineoplastic treatment comprises both l-O therapy and chemotherapy as defined above.

[0053] According to the invention, ILDR is defined as any low dose irradiation in the range of 0.25 to 4 Gy administered to the intestine or gut of a patient. In a preferred embodiment, the ILDR comprises or consists of exposing the patient’s intestines to at least 0.25 Gy and less than 4 Gy, preferably 0.5 to 3 Gy, more preferably about 1 Gy in one fraction. These fractions have been found to yield maximum abscopal effect while minimizing side effects.

[0054] In some embodiments, the ILDR is applied to the patient’s jejunum or ileum, i.e., the patient’s ileum or jejunum is irradiated. In some embodiments, the ILDR is applied to both the patient’s jejunum and ileum, i.e., the patient’s jejunum and ileum are both irradiated. In some embodiment, only the patient’s jejunum or the patient’s ileum is irradiated. In some embodiments, the ILDR is applied to the patient’s colon, i.e., the patient’s colon is irradiated, especially when irradiation of the patient’s jejunum and / or ileum is contraindicated. Preferably, the patient’s liver is not irradiated.

[0055] In some embodiments, the ILDR is administered to a volume of at most 1200, 900, 700, 600, 550, 500, 450 or 420 cc. In some embodiments, the ILDR is administered to a volume of at least 7, 15, 50, 70, 90 or 100 cc.

[0056] In some embodiments, the ILDR is concomitant with stereotactic ablative radiotherapy (SABR). As used herein, the term “stereotactic ablative radiotherapy (SABR)” refers to a highly focused radiation treatment where an intense dose of radiation is concentrated on a tumor.

[0057] In some embodiments, only one fraction of the ILDR is administered. In other embodiments, the fraction is repeated two to six times, preferably three times. In some embodiments, the fraction is repeated after 3 to 8 weeks. In other embodiments, the fraction is is repeated every other 6 weeks.

[0058] In some embodiments, the patient to be treated with the composition according to the invention has been diagnosed with a metastatic cancer.

[0059] In some embodiments, the patient has been diagnosed with a lung cancer, a colorectal cancer, a kidney cancer, a sarcoma, a pancreatic cancer, a melanoma, a cervical cancer or an anal cancer.

[0060] In some embodiments, the patient has been diagnosed with a cancer amenable to immunotherapy in first, second or third line therapy. In some embodiments, the patient’s tumor is refractory and / or does not respond to immunotherapy.

[0061] In some embodiments, the patient’s gut microbiota lacks bacteria belonging to the Christensenellaceae family, i.e., the patient does not have a detectable level of Christensenellaceae in his / her stools before administration of said composition.

[0062] In some embodiments, the patient’s gut microbiota lacks bacteria belonging to the Christensenella genus, i.e., the patient does not have a detectable level of Christensenella in his / her stools before administration of said composition.

[0063] In some embodiments, the patient’s gut microbiota lacks bacteria belonging to the Christensenella minuta species, i.e., the patient does not have a detectable level of Christensenella minuta in his / her stools before administration of said composition.

[0064] In some embodiments, the ILDR provokes an abscopal effect. As used herein, the term “abscopal effect” relates to a positive effect of a cancer treatment (e.g., inducing shrinkage) on one or more distant tumors that are not directly targeted by the treatment.

[0065] In some embodiments, the ILDR favors the growth of immunogenic species, e.g., of Christensenella minuta.

[0066] In some embodiments, the ILDR increases the patient’s response to the antineoplastic treatment.

[0067] In another aspect, the invention relates to a theranostic method for determining if a cancer patient needs a bacterial compensation before administration of a treatment combining ILDR and immunotherapy (l-O therapy), comprising assessing, in a feces sample from said patient, the presence of bacteria belonging to the Christensenellaceae family, wherein if no bacteria belonging to the Christensenellaceae family is present and / or detectable in said feces sample, the patient needs a bacterial compensation with a composition according to the invention.

[0068] The skilled person is well aware how the presence of bacteria of a certain family can be determined in a feces sample. For example, the bacteria can be detected by 16S rRNA sequencing or whole genome sequencing (WGS).

[0069] In another aspect, the invention relates to a theranostic method for determining if a cancer patient needs a bacterial compensation before administration of a treatment combining ILDR and immunotherapy (l-O therapy), comprising assessing, in a feces sample from said patient, the presence of bacteria belonging to the Christensenella genus, wherein if no bacteria belonging to the Christensenella genus is present / detectable in said feces sample, the patient needs a bacterial compensation with the composition according to the invention.

[0070] In yet another aspect, the invention relates to a theranostic method for determining if a cancer patient needs a bacterial compensation before administration of a treatment combining ILDR and immunotherapy, comprising assessing, in a feces sample from said patient, the presence of Christensenella minuta, wherein if no Christensenella minuta is present / detectable in said feces sample, the patient needs a bacterial compensation with the composition according to the invention.

[0071] In the theranostic methods according to the invention, the treatment combining ILDR and immunotherapy (l-O therapy) may be one as described above in relation with the compositions according to the invention.

[0072] In the theranostic methods according to the invention, the patient may be one as described above. In particular, the patient may be diagnosed with a lung cancer, a colorectal cancer, a kidney cancer, a sarcoma, a pancreatic cancer, a melanoma, a cervical cancer or an anal cancer. More particularly, the patient has been diagnosed with a cancer amenable to immunotherapy in first, second or third line therapy. In some embodiments, the patient’s tumor is refractory and / or does not respond to immunotherapy.

[0073] In another aspect, the invention relates to a method of treating cancer in a patient in need thereof, comprising administering an effective amount of the composition according to the invention, in combination with ILDR and an antineoplastic treatment. That is, the invention relates to method of treating cancer in a patient in need thereof, comprising administering an effective amount of the composition live bacteria belonging to the Christensenellaceae family (e.g., a LBP comprising Christensenella minuta 601 (deposited as CNCM 1-6155), in combination with ILDR and an antineoplastic treatment.

[0074] In the method of treatment according to the invention, the ILDR may be one as described above in relation with the compositions according to the invention.

[0075] In the method of treatment according to the invention, the antineoplastic treatment may be one as described above in relation with the compositions according to the invention.

[0076] In the method of treatment according to the invention, the patient and / or the cancer may be as described above in relation with the compositions acording to the invention. In some embodiments, the patient is a patient identified as in need of a bacterial compensation by the theranostic methods according to the invention.

[0077] In some embodiments, the method of treatment according to the invention is for or is capable of increasing or restoring the patient’s response to the antineoplastic treatment.

[0078] In another aspect, the invention relatest to a method of treating cancer in a patient in need thereof, comprising the following steps:

[0079] (i) optionally, assessing the presence of (a) bacteria belonging to the Christensenellaceae family, (b) bacteria belonging to the Christensenella genus or (c) Christensenella minuta in a feces sample from said patient;

[0080] (ii) optionally, especially if step (i) has been performed and has showed that the searched bacteria are not present, administering an effective amount of a composition according to the invention;

[0081] (iii) performing ILDR as described above;

[0082] (iv) administering an antineoplastic treatment as described above; wherein step (iv) is performed between one day to three weeks, preferably less that 5 days after step (iii).

[0083] Preferably, in the method of treating cancer according to the invention, step (iv) comprises administering PD1 blockade (e.g., an anti-PD1 or anti-PD-L1 Ab). In some embodiments, step (iii) is performed after a first injection of anti- PD1 or anti-PD-L1 Ab, for example three weeks after said first injection.

[0084] In some embodiments, the treatment with said anti-PD1 or anti-PD-L1 Ab is pursued every other 3 weeks for up to 18 months or more until disease progression.

[0085] In some embodiments, the patient has been diagnosed with a cancer amenable to immunotherapy in 1st, 2ndor 3rdline therapy.

[0086] In some embodiments, said patient’s response to said immunotherapy is insufficient before ILDR.

[0087] Preferably, the method comprises several cycles of ILDR + PD1 blockade.

[0088] In some embodiments of the method of treating cancer, the composition according to the invention is administered at least in one cycle. Preferably, the composition according to the invention is administered in each cycle.

[0089] EXAMPLES

[0090] Example 1 : Inflammatory, metabolic and metagenomic profiling of responders to SABR-aPD-L1

[0091] We sought to identify gut microbiota-associated biomarkers that could segregate Rs from NR patients based on shotgun MGS of 48 fecal samples (29 at V1, 13 at V3 and 6 at V4) (Chen et al 2025). . Members of the Eubacteriales order (Ruminococcus bromii SBG4285, Eubacterium rectale SGB4933, Ruminococcaceae bacterium SGB14925) were relatively over-represented in Rs (considering both Elite patients or OS12), while the Enterocloster genus (E. aldensis SGB4762, E. bolteae SGB4758) was relatively dominant in NRs (Figure 1A, Figure 6, Tables 1 and 2). Of note, Parabacteroides distasonis SGB1934, already reported as a species relatively enriched in chemoradiotherapy-resistant rectal cancers (Teng et al, 2023), stood out in non-elite patients (Figure 1A, Tables 1 and 2). Interestingly, we observed the relative overabundance of Christensenellaceae family members (C. minuta, Christensenellaceae bacterium SGB14127) and their associated archaea Methanobrevibacter smithii (Ruaud et al, 2020) in the Elite and OS12 patients respectively (Figure 1A, Figure 6). Exploiting the MG data base of 5234 healthy volunteers (HV) (Pasolli et al, 2017), we confirmed that the aforementioned bacterial species prototypically associated with Rs (C. minuta and R. bromii) were underabundant in NRs compared with HV and, vice versa, NR-associated species (P. distasonis and E. aldensis) were overabundant in NRs (Figure 1B).

[0092] Example 2: ILDR administered at 1 Gy but not 4 Gy is associated with clinical benefit to PD-L1 blockade in mice

[0093] The positive impact of ILDR on patient long-term survival and response to RT appeared intriguing. To eliminate the potential impact of RT-induced immunogenic cell death of cancer cells in this effect, we took advantage of a murine ILDR model to evaluate its direct capacity to boost the immunostimulatory effects of ICB against a heterotopically transplanted colon cancer (MC38) inoculated distally, outside of the bowels and the field of irradiation. The experimental setting is illustrated in Figure 2A including a representative image of the irradiation beam for 1Gy ILDR on the right. Hence, the subcutaneous tumor was spared from all RT beams to strictly evaluate the contribution of ILDR ± aPD-L1 Ab to tumor shrinkage. Small and large intestines could not be distinguished from each other on CT images; therefore, both were encompassed in the field. As expected, aPD-L1 alone resulted in tumor growth retardation or transient regression, but all tumors irreversibly relapsed (Figure 2B). Animals subjected to ILDR alone displayed no survival benefit compared with controls (median 24 vs 26 days, P > 0.05). Of note, the combination of ILDR and aPD-L1 Ab mediated synergistic antitumor effects, culminating with the cure of 30% of the animals. No overt signs of adverse events (i.e. , weight loss or bloody feces) were observed in mice subjected to ILDR, alone or in combination with aPD-L1 Ab. We next evaluated the antitumor effects of PD-L1 blockade alone or together with different ILDR doses. While 1Gy of ILDR improved CR rates (up to 30%) with aPD-L1 , lower (0.25Gy) or higher (4Gy) ILDR dosing failed to do so (0% tumor free mice respectively), suggesting a narrow dose window to achieve such a synergy (Figure 2C). ILDR delivered onto the liver was not effective (Figure 2D). The beneficial role of incidental irradiation outside the tumor field was also observed in a spontaneous breast cancer model induced by administration of medroxyprogesterone acetate and 7,12-dimethylbenz[a]anthracene in C57BL / 6J mice, as previously described (Buque et al, 2020) (Figure 7). A total of 37 mice with primary breast tumor developing either in the cervical (n = 22) or abdominal (n = 15) mammary fat pad were subjected to mock (n = 18) or SABR (20Gy x 2F or 20Gy x 3F, n = 19) delivered to the first primary tumor (Figure 7A, left). A post-hoc analysis was performed on the survival data of this preclinical study. The intestinal dosimetry data were not available. Mice with abdominal SABR were compared with those subjected to cervical SABR (Figure 7A, right). In response to mock SABR, both groups displayed comparable OS and time to secondary (TTS) tumor, a surrogate marker of abscopal effects (Figure 7B). However, delivery of SABR onto abdominal lesions (but not cervical ones) significantly improved OS (median 60 days vs 41 days, P = 0.0285) as well as TTS (median 48 days vs 5 days, P = 0.0116) (Figure 7C).

[0094] In two experimental tumor models, ILDR appeared to boost the immunostimulating activity of PD-L1 blockade, in accordance with epidemiological associations observed in patients.

[0095] Example 3: Favorable and deleterious role of Christensenellaceae and Bacteroidales in the ILDR immunostimulatina effects respectively

[0096] Given the associations between the taxonomic composition of feces and the clinical outcome, we analyzed the impact of the gut microbiota on the antitumor efficacy of the combination of ILDR+ PD-L1 blockade. We selected 8 independent patient stool samples (fecal microbial transplants (FMT)) that we transferred by oral gavage into avatar mice that were then inoculated with sc MC38 and treated according to the protocol aligned in Figure 2A. One FMT out of two could lead to complete tumor eradication, suggesting that the stool composition at baseline mattered for the success of ILDR during immunostimulation (Figure 2E-F).

[0097] We next assessed the kinetics of the abundance of indigenous bacterial species in murine feces over the course of the combinatorial regimen, by 16S rRNA sequencing followed by qPCR primer sets specific of distinct commensal species identified fortheir potential relevance in patients (Figure 6). Firstly, we used the unbiased approach based on 16S rRNA amplicon sequencing of stools in a longitudinal manner to analyze potential significant differences in the kinetics of distinct species favored by 1Gy ILDR in the setting of PD-L1 blockade (Table 3). Comparisons of beta-diversities in principal component analysis (PCoA) between V2 and V4 as well as V2 and V5 (refer to Figure 2A for the time points) within each mouse group (1Gy ILDR + aPD-L1) showed significant differences (Table 3, Anosim: P =0.32, Permanova: P =0.02 for V2 vs V4, Anosim: P =0.11 , Permanova: P =0.0008 for V2 vs V5). Moreover, the PCoA was significantly different between the two groups (1Gy ILDR + aPD-L1) versus (4Gy ILDR + aPD-L1) at V4 (Anosim: P =0.006, Permanova: P =0.003 at V4) but not at V5 (Table 3). In particular, several species selected from the two groups 1Gy versus 4 Gy ILDR + aPD- L1 did not exhibit the same kinetics of relative representativity. The addition of 1Gy ILDR but not 4Gy ILDR, to aPD-L1 appeared to favor the growth of immunogenic species, including the health-related Christensenella massiliensis , the xylanolytic Lacrimispora aerotolerans and Odoribacter splanchnicus, which confers protection from colitis and CRC in mice and improves longevity . High-dose ILDR failed to boost the abundance of R species resulting from aPD-L1 Ab ± 1Gy ILDR, such as the cellulolytic Ruminococcus flavefaciens . Conversely, 4Gy ILDR resulted in overrepresentation of taxa endowed with tolerogenic potential, particularly the BA deconjugating species Muribaculum intestinale and Culicoidibacter larvae (Figure S4 of Chen et al, 2025). Again, using the PCR method, we confirmed that PD-L1 blockade did not significant affect the prevalence or relative abundance of these murine species, Moreover, we confirmed that 1Gy ILDR led to increased relative abundance of two Rs-associated species (E. rectale and R. bromii) (Figure 8A), but not the NRs-related immunosuppressive species of the Enterocloster genus (E. clostridioformis), as observed in patients (Figure 8B). Of note, the augmentation of Rs species was absent or marginal in mice subjected to 4Gy ILDR.

[0098] Since the efficacy of ILDR + aPD-L1 might rely on health-related commensals such as Christensenellaceae family members in mice and patients, we analyzed the impact of several species and strains of various commensals available in our laboratory orally gavaged prior to ILDR and PD-L1 blockade in MC38 tumor bearers. Surprisingly, out of seven strains, only Christensenella minuta strains and especially one, i.e Christensenella minuta 601 , significantly boosted the curative efficacy of the ILDR+aPD-L1 Ab regimen (Figure 2G-H). C. minuta 601 was cultivated from the healthy ileum of a patient diagnosed with an ileal adenocarcinoma, a very rare tumor location in the digestive tract. Actually, the other strains appear to inhibit the ILDR effect, and belonged to the Bacteroidales order that has been shown to attenuate radiotherapy- induced DNA damage by suppling pivotal nucleotides counteracting DNA damage (Teng 2023). Indeed, we next confirmed that N Rs-associated species could blunt the antitumor efficacy of the combinatorial regimen. We focused on Parabacteroides distasonis, a major NRs species that tended to exhibit a higher abundance in patients with shorter OS in the SABR trial (Figure 9A, P =0.06). In these patients, the abundance of P. distasonis significantly correlated with the guanosine biosynthesis pathway but anticorrelated with purine-pyrimidine degradation, (Figure 9B). We next determined whether the presence of P. distasonis in the gut affected the synergy between ILDR and aPD-L1 using the MC38 model. P. distasonis was not found in the syngeneic C57BL / 6J mice from our suppliers. Therefore, we colonized these mice with P. distasonis by oral gavage of a strain of this species isolated from the ileal mucosa of a colon cancer patient. The efficacy of P. distasonis colonization was validated by qPCR (Figure 9C, left). Oral gavage with live P. distasonis dramatically did not impact control tumor growth (Figure 9C, middle) but reduced the synergistic effect of the combination of 1Gy ILDR and aPD-L1 (Figure 2I, right, Figure 9C, right, P = 0.03), as it triggered a rise in plasma deoxyguanosine concentrations (Figure 9D) and led to increased infiltration of a4p7+ regulatory T cells in the tumor bed (Figure 9E). These results suggest that the presence of distinct NRs-related commensals belonging to the Bacteroidales order resident in the gut prior to RT may compromise the potential synergy between ILDR and PD-L1 blockade.

[0099] Altogether, the gut microbiota composition governs the synergistic effects between ILDR and PD-L1 blockade in mice, with species from the Christensenellaceae family associated with favorable outcome in contrast to species from the Bacteroidales order.

[0100] Example 4: 1Gv ILDR enhanced the therapeutic efficacy of aPD-L1 Ab by promotinq qut emiqration of PD-L1 expressinq intestinal DC and effector CD8+T cells

[0101] To dissect the immune mechanisms underlying the efficacy of the combinatorial regimen, we depleted CD4+or CD8+T cells in mice using specific monoclonal antibodies. Depletion of CD8+, but not CD4+T cells, abrogated the synergistic antitumor effects of 1Gy ILDR + aPD-L1 Ab, suggesting that CD8+T cells are the key effectors (Figure 3A). IL-7, the elite-related cytokine known to maintain lymphocyte survival and to block TGFb immunosuppressive effects was undetectable in mouse plasma samples in the combinatorial regimen. In vitro irradiation of murine ileal organoids stimulated IL-7 secretion by 48 hours in a dose-dependent manner (Figure 9E). In contrast, in the context of PD-L1 blockade in vivo, ileal tissues from mice subjected to 4Gy ILDR + aPD-L1 Ab tended to display lower H7 transcripts compared to those from mice treated with aPD-L1 Ab alone (Figure 9F). Next, we blocked IL-7R using specific neutralizing antibodies but failed to interfere with the ILDR+anti-PD-L1 Ab curative effects (Figure 9G). We next immunophenotyped tumor-infiltrating lymphocytes (TILs) from mice treated with isotype Ab or aPD-L1 Ab with or without 1Gy or 4Gy ILDR at day 11. TCF1 and TOX are two transcription factors influencing T cell differentiation and ICB treatment efficacy. 1 Gy ILDR slightly increased infiltration of MC38 tumors by CD8+T cells (Figure 3B) while 4 Gy boosted Treg proportions (Figure 10A). Moreover, while 1 Gy ILDR boosted the PD1+CXCR6+memory CD8+T cell subset, 4 Gy increased the TOX1 expressing CD8+T cells (Figure 3C, Figure 10C) as compared to isotype control or aPD-L1 Ab, in line with a recent report that high-dose irradiation (16.0-18.4Gy) of gut and kidney healthy tissue upregulated Tox transcripts in TILs. Moreover, the systemic recirculation of enterotropic a4b7+CD4+Foxp3+regulatory T cells (Tregs) and RORyt+Tregs (Tr17) cells was reduced after 1Gy (but not as much after 4 Gy ILDR), allowing an increase of the CD8 / Treg ratio of intestinal T cell emigrants in the tumor immune infiltrate (Figure 10D, Figure 3D-E).

[0102] There is growing evidence that tumor responses to ICB require intestinal DC emigration to mesenteric lymph nodes that mediate cross-presentation of bacterial or tumor antigen to specific T cells. Ionizing radiation (IR) may enhance the immunogenicity of DC but direct high-dose (10-30Gy) irradiation of DCs impaired their antigen-presenting cell (APC) function. We undertook to monitor the quantitative and qualitative dynamic fluctuations of intestinal DC in mesenteric and skin tumor draining lymph nodes by flow cytometry. First, we observed that percentage of mregDC (CD45+ lineage negative (CD3 19-NK1.1-Ly6G’Siglect F’) CD11c+MHC ll+, PD-L1+CD40+) in DC or CD45+cells increased in TdLN and mesenteric lymph nodes (mLN) to various extents following ILDR (Figure 3F, Figure 11). Next, we followed intestinal DC exodus from the gut, tracking the activated DC subsets from mLN to TdLN using injection of CFSE into mLN (Figure 3G). Indeed, the influx of PD-L1+DC into TdLN was originating from the gut (Figure 3G, right panel). Adding C. minuta by means of oral gavage dramatically increased the accumulation of activated DC expressing CCR7 into TdLN (Figure 3H).

[0103] To dive into the potential mechanisms accounting for mLN DC activation, we performed a targeted plasma metabolomics of tumor bearers subjected to ILDR+aPD-L1 Ab. We first performed targeted metabolomics of plasma collected at day 11 (3 days after ILDR) from mice treated with the combinatorial regimen (Figure 3I). The comparison between plasma post-aPD-L1 Ab versus aPD-L1 Ab + 1Gy ILDR revealed that the combinatorial regimen significantly affected the balance between butyrate- related metabolites (2-(OH)-3-methylbutyric and isobutyric acids) and lipid derivatives such as cholesterol and alpha-tocopherol (Figure 3I). Indeed, 1Gy ILDR markedly reduced the plasma level of butyrates as compared to aPD-L1 or isotype Ab (Figure 3I, left panels). In contrast, the abundance of both cholesterol and alpha-tocopherol declined in mice subjected to either aPD-L1 Ab or 1Gy ILDR alone, but was fully restored upon combination treatment (Figure 3I, right panels). These data are in line with recent findings showing that butyrate compromises the RT-mediated anti-tumor immune responses, while cholesterol and alpha-tocopherol act on the dendritic cell (DC) / T cell cross-talk, thereby enhancing anticancer immune responses upon ICB. Using a DC-T cell co-culture system that was previously described (Zhao et al 2021 , 2023), we examined the dose (0-6Gy) effects of IR onto the cross-presentation capacity of DC loaded with ovalbumin (OVA) protein, monitored by the reactivity (IL-2 release) of an O A257-264-specific, H-2Kb-restricted cytotoxic T lymphocyte (CTL) hybridoma (B3Z) (Figure 3J). Of note, doses that appeared optimal in ILDR (i.e., 1 or 3Gy) significantly augmented IL-2 production in this system measuring DC function, but IR doses outside of this range (i.e., 0.25, 4 or 6 Gy) failed to do so (Figure 3K). Next, we used the same experimental system to screen a compound library comprising 96 BAs as well as metabolites relevant to carnitine and sterol metabolism. Strikingly, cholesterol and distinct bile acids stood out among the few metabolites that significantly augmented APC function of DCs (Figure 3L). Altogether, these in vitro findings suggest that ILDR could mediate its CD8+T cell effector-stimulatory effects via direct and indirect, metabolic effects on DCs.

[0104] Example 5: ILDR reduced lactate accumulation while increasing immunogenic indole and secondary bile acid metabolites in plasma

[0105] The mass spectrometry-based targeted plasma metabolomics of tumor bearers subjected to ILDR+aPD-L1 Ab also revealed a drop in lactic acid -related metabolites (Figure 4A-B), in accordance with the reported deleterious effects of L-lactate produced by tumoral Lactobacilli species (Colbert et al, 2023). We next sought for correlations between the various Christensenella species detectable in murine and patient stools and all the >200 plasma metabolites that were monitored. In mice, the relative abundance and prevalence of each of the four detectable Christensenella species were variable (Figure 4C) and the most prevalent C. massiliensis was significantly associated with secondary bile acids (BA) (deoxycholic acid (DCA) and taurolithocholic acid) (Figure 4D). Using a DC-T cell co-culture system (Zhao et al 2021 , 2023), we examined the dose effects of various BA including several secondary BA onto the cross-presentation capacity of DC loaded with ovalbumin (OVA) protein, but only DCA and UDCA (ursoDCA) could significantly increase DC cross-presentation (Figure 4E). Importantly, in patient stools and plasma again, we observed a 30% fecal prevalence of C. minuta, that was stable over the course of the SABR protocol (Figure 4F) and a significant correlation with the secondary BA taurodeoxycholic acid (Figure 4G). In addition, as previously described, the immunogenic arginine and indole-3 propionate and indole-3 acetate were also standing out in those patients harboring intestinal C. minuta.

[0106] Taken together, these findings indicate that 1Gy ILDR may boost the immunostimulatory effects of PD-L1 blockade by favoring the relative growth of Christensenella species implicated in the exodus of PD-L1+mregDC from the mLN to TdLN and the circulation of immunogenic plasma metabolites (secondary biliary salts, indoles, arginine) while limiting anaerobic glycolysis.

[0107] Example 6: Phase 2 ILDR study in cancer patients who were refractory to PD1 / L-1 blockade

[0108] The stage I of a trial entitled ‘Intestinal low dose radiotherapy combined with immunotherapy in immune-resistant cancer patients diagnosed with metastatic Solid Tumors’ will recruit 16 patients for 1Gy ILDR, while 44 patients will be enrolled in stage II for ILDR dose escalation. The primary objective of the study is to evaluate the efficacy and safety of ILDR combined with PD-1 inhibitors in immune-resistant metastatic malignant solid tumors. Occurrence rates of disease control rate (DCR), time to second objective disease progression (PFS2), ORR, and adverse events in accordance with the guidelines for response criteria for use in trials testing immunotherapeutics (both RECIST and iRECIST) will determine the extent of benefit for patients participating in this trial. At least one accessible and measurable lesion should be selected as the target lesion for observation. The jejunum and ileum will be selected and performed with low dose radiation of 1Gy / F, 1Gy-3Gy within 10 days (Figure 5A). The primary immunotherapy regimen that each patient has developed resistance to will be reintroduced in combination with radiotherapy, with a treatment frequency of once every 3 weeks until disease progression, patient death, or unacceptable toxicity. Secondary objectives are the fluctuations in metagenomics, metabolomics, and immunomics in these three groups.

[0109] We report on the clinical results of the first 10 patients (Clinical characteristics in Table 4) consecutively enrolled in the stage I who met the eligibility criteria after obtaining their written informed consent. All patients completed 1Gy ILDR and at least 1 cycle of ICB. The majority of them displayed Grade 1-2 toxicities during treatment. Two patients (20%) experienced Grade 3-4 liver toxicities and had recovered after ceasing ICB. A representative dosimetric graph of 1Gy ILDR is shown in Figure 5B. The waterfall graph of the best tumor response (by RECIST 1.0) of all 10 patients is depicted in Figure 5C, showing two partial responses, 5 stable diseases and 3 progressions. Of note, the first PR case started with a SD for 2 months, transformed into a PR durable over 8 months. The second PR case only had 4 months follow up, meaning that a confirmation is pending at 6 months. Figure 5D and 5E highlight the spiderplots depicting the evolution of the target lesion and the swimming plot of each patient outcome. A representative CT scan of a scalp lesion before and after 3 months of combinatorial therapy is shown (Figure 5F). Another patient presenting with thoracic lymph node regressions is depicted in Figure 5G.

[0110] We conclude that ILDR and immunotherapy may have potential to circumvent resistance to PD1 blockade in advanced cancer patients.

[0111] Discussion

[0112] Patients who have simultaneous abdominal and extra-abdominal lesions might gain more survival benefits from SABR when it is delivered to abdominal tumor deposits and hence causes ILDR as an 'accidental' byproduct. Importantly, when treating purportedly abdominal metastases, one might consider to deliver a selected dosing within the > 1 and <3 Gy range. Such an additional ILDR is unlikely to aggravate patient toxicities, as suggested by the SABR trial, the Phase II pilot study reporting the first 10 patients as well as by preclinical model experimentation. Moreover, microbial factors affect the probability of ILDR to yield positive effects. This is suggested by the predominant gut dysbiosis in SABR non-elite patients, as well as the detrimental effect of P. distasonis and bacteria of the Bacteroidales order in general, on the synergy between ILDR and aPD-L1 Ab. In contrast, Christensenellaceae family members (such as C. minuta, specifically strain 601 , and potentially species C. massiliensis) were associated with Elite patients, benefited from ILDR for their intestinal relative dominance, and markedly increased curative effects of the synergy between ILDR+aPDL1 Abs in mice after oral gavage. Assessment of the gut microbiota prior to treatment might help select patients that benefit from ILDR in the context of PD-1 / PD-L1 blockade. Patients suffering from dysbiosis might require additional interventions to correct their microbiota, such as fecal material transplantation which reportedly improves tumor response to ICB.

[0113] CD8+T cells proved to mediate the beneficial effect of ILDR combined with aPD-L1 Ab. PD1 / PD-L1 blockade alone can drive the proliferation and differentiation of stem-cell like CD8+T cells, resulting in the expansion of exhausted effector cells in mice. Addition of ILDR to aPD-L1 resulted in a substantial increase in CD8+T cell infiltration of the tumor bed. 1Gy ILDR increased CXCR6 expression in TCFT CD44+PD1+exhausted effector T cells, which is critical for their survival in the tumor microenvironment. 1Gy (but not 4 Gy) ILDR also reduced the homing of enterotropic Tregs to tumor lesions, which should boost local tumor immunosurveillance. In contrast, 4 Gy (but not 1Gy) ILDR was associated with increased T cell exhaustion defined by the upregulation of TOX in TCFTCD44+PD1+CD8 TILs, indicating a transition from a proinflammatory to an immunosuppressive tumor landscape. Consistently, Tadepalli et al. showed that incidental high-dose (16.0-18.4Gy) irradiation of gut and kidney augmented TOX expression in TILs, accompanied by rerouting of proinflammatory monocytes from the tumor to colon. Lastly, in contrast to 1 Gy, 4 Gy failed to activate the DC / T cell cross-talk in vitro, stressing the dose-dependency of this effect. Importantly, CFSE injection allowed to visualize the CCR7+ PD-L1+ DC trafficking from mLN to tdLN, exacerbated by C. minuta, and unleashed by PD-L1 blockade. The relevance of the emigration of mLN DC to tdLN has been elegantly demonstrated in a previous report in the setting of CTLA4+PD1 coblockade.

[0114] The data presented suggest that ILDR could mediate distal immunostimulation, most probably through the modulation of the microbiota-host interaction leading to a shift in the systemic metabolism. Specifically, ILDR affected cholesterol and BA metabolism, increasing blood levels of conjugated secondary BAs, which were found to augment DC APC function. Cholesterol accumulation in DCs may enhance antigen presentation and T priming, mediating autoimmune diseases. In the scenario of metastatic cancer, this proinflammatory effect might help improve tumor responses to ICB. In contrast, NRs patients harboring increased abundance of lipids, fatty acids, and their oxidized products may be resistant to the immunostimulatory effects of ILDR on DC. Similarly, the short chain fatty acid butyrate produced by the intestinal ecosystem reportedly inhibits the stimulator of IFN genes pathway in DC, dampening CTL-mediated immunity in response to RT, as observed in our mouse model.

[0115] Altogether, our findings indicate that ILDR might enhance the anti-tumor CD8+T cell response by improving the metabolic fitness of intestinal DCs, hence adding another level of complexity to the regulation of the gut-tumor immune axis.

[0116] Previous studies investigating the effects of IR on gut microbiota composition described a reduction of microbiota richness and an increase in the relative abundance of pathobionts (such as Proteobacteria, Bacteroidetes, Fusobacteria) over the health-related Faecalibacterium and Bifidobacterium in patients (Teng et al 2023, Fernandes et al, 2021 , 2023; Liu et al, 2019). Our 1Gy-ILDR microbiome fingerprint differs from these reports, highlighting the relative over-representation of health-related immunostimulatory commensals (so called “R species”) including Christensenellaceae spp, while failing to affect harmful pathobionts (such as species of the Enterocloster genus, Bacteroidales order) in tumor bearers and in Elite or OS12 Rs patients. Indeed, the Enterocloster genus, associated with resistance to SABR in our patients can cause significant perturbations of the BA composition, leading to the exodus of enterotropic Tr17 cells to tumor beds and tumor progression (Fidelle et al, 2023), while conjugated secondary BAs may be important for the DC / T cell cross-talk. In addition, specific pathobionts (such as Parabacteroides spp) may subvert the effects of RT during rectal cancer chemoradiotherapy through the stimulation of purine / pyrimidine biosynthesis and a consequent reduction of RT-induced DNA damage (Teng et al 2023).

[0117] In sum, our study suggests that the combination of SABR and aPD-L1 Ab is well tolerated in advanced cancer patients with enhanced-therapeutic effects in a subset of patients who received off- target ILDR. The effect is dependent on exposure of off-target gut to a mean radiation dose of 1Gy (and it is lost at mean dose of > 4Gy), and on the baseline host-microbiota interactions compatible with intestinal PD-L1+ DC emigration and CD8+T cell activation.

[0118] Experimental model

[0119] Animals

[0120] Female, 6-7 weeks old, wild type (WT) C57BL / 6J mice were purchased from Envigo (France) and housed in specific pathogen-free conditions at the animal facility of Gustave Roussy in France. In indicated experiment, WT C57BL / 6J female mice of 6-7 weeks of age were obtained from The Jackson Laboratory and housed under standard conditions at Weill Cornell Medical College (New York, NY) in line with institutional requirements, on a 12 light / 12 dark light cycle, 20-25 °C, and 30% humidity.

[0121] Cell lines

[0122] MC38 cell line (syngeneic from C57BL / 6J mice) were purchased from Sigma and cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS), 2mM L-glutamine, 100 III ml’1penicillin / streptomycin, 1mM sodium pyruvate and nonessential amino acids. Mycoplasma test in cell culture medium was routinely performed and all tests were negative. Cells were used within passage number 8.

[0123] The inducible immortalized dendritic cells (iniDCs) were a gift from Cornelia Richter and colleagues (Richter et al., 2013), and cultured in basic DC medium: RPMI-1640 medium supplemented with 10% decomplemented FBS, 1 mM sodium pyruvate, 10 mM HEPES, 1X penicillin / streptomycin, 50 pM p-Mercaptoethanol and 10ng / ml recombinant GM-CSF. IniDCs cells were immortalized under the induction of Dex / Dox (dexamethasone at 100 nM + doxycycline hyclate at 2 pM) and differentiated into immature DCs (“de-iniDCs”) upon removal of Dex / Dox.

[0124] The B3Z hybridoma T cells were kindly provided by Sebastian Amigorena and maintained with DC medium supplemented with p-mercaptoethanol (50 pM).

[0125] Murine intestinal organoids

[0126] Crypt-derived organoids were generated from ileal tissues of 8-10 weeks old C57BL / 6J female mice as described previously (Roberti et al, 2020). They were cultured in advanced DMEM / F12 medium supplemented with 100 II ml’1penicillin G sodium, 100 pg ml’1streptomycin sulfate, 2mM L-glutamine, 10 mM HEPES, 1X N2 supplement, 1X B27 supplement, 50ng ml’1mEGF, 100mg ml’1mNoggin (Petrotech), N-acetylcysteine (Sigma) (reagents from Invitrogen unless otherwise indicated) and 10% conditional medium of HA-R-Spondin1-Fc 293T Cells.

[0127] Method details

[0128] Metagenomic sequencing of fecal samples

[0129] Fecal samples were collected from patients at multiple time points (V1 , before the 1stdose of aPD-L1 Ab; V3, after SABR and before the 3rddose of aPD-L1 ; V4, at 6 months post the 1stdose of aPD-L1) following the International Human Microbiome Standards (IHMS) guidelines. The stools were subjected to total DNA extraction and sequencing with Ion Torrent technology following MetaGenoPolis (INRAE), France as previously reported (Routy et al, 2018; Fidelle et al, 2023). Microbiome taxonomic and functional composition of each sample have been estimated via MetaPhlAn-4 (Blanco-Miguez et al, 2023) and HUMAnN-3 (Beghini et al, 2021) profiling tools, respectively. Shannon index was calculated by applying the diversity function available in the Vegan R package. PERMANOVA analysis was performed applying the adonis2 function (Vegan R package) with default parameters and principal coordinate decomposition was computed via the pcoa function available in the ape R package. We used the Mann-Whitney II test to estimate differential abundant microbiome features for each of the considered conditions. Microbiome features that were present in less than 5 samples in one of the two tested sets were excluded from the analysis. Adjusted P-values (Q) are computed via the Benjamini-Hochberg procedure to control for false discovery rate (FDR). All the reported associations for differential abundance analysis presented P < 0.05 and Q > 0.2. Spearman correlation analysis was performed to detect associated microbial pathways involved in nucleotide metabolism with Parabacteroides distasonis SGB1934.

[0130] 16S rRNA-Seq and analysis of murine fecal samples

[0131] Fecal samples were collected from individual mice enrolled in indicated experiments and subjected to DNA extraction using the QIAamp Fast DNA Stool Mini Kit (QIAGEN) according to the manufacture’s protocol. Fecal DNA samples were sequenced at IHU Mediterranee Infection, targeting the V3-V4 regions of the 16S rRNA gene with Illumina MiSeq technology.

[0132] Raw FASTQ files were analyzed with DADA2 pipeline v.1.14 (Callahan et al, 2016) for quality check and filtering on a Workstation Fujitsu Celsius R940 (Fujitsu,

[0133] Tokyo, Japan). Raw reads (12460891 in total, on average 78866 per sample) were filtered (2377392 in total, on average 15047 per sample) and 2849 Amplicon Sequence Variants (ASV) were found with a 99% coverage for all samples. Bioinformatic and statistical analyses on recognized ASV were performed with Python v.3.8.2 (https: / / www.python.org / ). Each ASV sequence underwent a nucleotide Blast using the National Center for Biotechnology Information (NCBI) Blast software (ncbi-blast-2.3.0) and the latest NCBI 16 S Microbial Database (ftp: / / ftp, ncbi.nlm.nih.gov / blast / db / ). After blasting, the 2849 ASVs were merged into 392 species. Of them, 142 species with > 20% prevalence were considered for subsequent statistical analyses.

[0134] Data matrices were first transformed with pseudocount and centered-log- ratio (CLR), then normalized and standardized using QuantileTransformer and StandardScaler methods from Sci-Kit learn package v1.0.1 (https: / / scikit- Measurements of a diversity was calculated at species level using the

[0135] SciKit-learn package v1.0.1. Exploratory analysis of p-diversity was calculated using the Bray-Curtis measure of dissimilarity and represented in Principal Coordinate Analyses (PCoA), along with methods to compare groups of multivariate sample units (analysis of similarities - ANOSIM, permutational multivariate analysis of variance - PERMANOVA) to assess significance in data points clustering. PCoA and the other analysis was performed on batch-effect corrected data with MMUPHin (Ma et al, 2022) . Batch- corrected data were assessed by two-way interaction terms in the Mixed Model with Principal Variance Component Analysis (PVCA) and visualized with Uniform Manifold Approximation and Projection (UMAP), both implemented within BatchServer (https: / / lifeinfor.shinyapps.io / batchserver / ). In order to visualize a non-supervised clusterization as PCoA, we implemented with custom scripts (Python v3.8.2, Seaborn vO.11.2) a Hierarchical Clustering Analysis (HCA) with ‘Bray-Curtis’ metrics and ‘complete linkage’ method. We implemented Partial Least Square Discriminant Analysis (PLS-DA) and the subsequent Variable Importance Plot (VIP) as a supervised analysis wherein the VIP values (order of magnitude) are used to identify the most discriminant bacterial species among the cohorts. Mann-Whitney II test and P values, without FDR, was used for a fixed sample size as previously described and Kruskal-Wallis tests were employed to assess significance for pairwise or multiple comparisons, respectively, considering a P value < 0.05 as significant. All P values were corrected for multiple hypothesis testing using a two-stage Benjamini-Hochberg FDR at 10%. Lineplot temporal kinetics of bacterial species relative abundances were computed for each timepoint and depicted with Seaborn vO.11.2, representing data variance with 95% confidence intervals.

[0136] Metabolomic analyses of serum / plasma samples

[0137] Samples were treated following the protocol described previously (Grajeda-Iglesias et al, 2021) to extract metabolites. Briefly, 50 pL of serum were vortexed with 500 pL of ice-cold extraction mixture (MeOH / water, 9 / 1 , -20°C, with a cocktail of internal standards) during 5 minutes, and then centrifuged (10 min at 15000 g, 4°C). Several fractions were split to be analyzed by different Liquid and Gaz chromatographies coupled with mass spectrometers (LC / MS and GC / MS) as previously described (Durand et al, 2021). Widely targeted analysis by GC-MS / MS was performed on a 7890A gas chromatography (Agilent Technologies) coupled to a QQQ (triple quadrupole) 7000C (Agilent Technologies). Polyamines, short chain fatty acids and bile acids analyses were performed by LC-MS / MS with a 1290 UHPLC (Ultra-High Performance Liquid Chromatography) (Agilent Technologies) coupled to a QQQ 6470 (Agilent Technologies). Widely pseudo-targeted analysis by UHPLC-HRMS (Ultra-High Performance Liquid Chromatography - High Resolution Mass Spectrometer) was performed on a Dionex U3000 I Orbitrap q-Exactive (Thermo) coupling. All targeted treated data were merged and cleaned with a dedicated R (version 4.0) package (@Github / Kroemerlab / GRMeta).

[0138] Multiplex enzyme-linked immunosorbent assay (ELISA)

[0139] Serum samples were prospectively collected from patients enrolled in the SABR trial at multiple time points (V1 , before the 1stdose of aPD-L1 Ab; V2, before the 2nddose of aPD-L1 and SABR; V3, after SABR and before the 3rddose of aPD-L1 ; V4, at 6 months post the 1stdose of aPD-L1). The concentration of 27 cytokines / growth factors (IL-1 p, IL-1 Ra, IL-2, IL-4, IL-5, IL-6, IL-7, IL-8, IL-9, IL-10, IL-12, IL-13, IL-15, IL- 17, Eotaxin, FGF basic, G-CSF, GM-CSF, IFNy, IP-10, MCP-1 , MIP-1a, PDGF, MIP-1 , RANTEs, TNFa, VEGF) in these samples was assessed using a Bio-Plex Pro Human Cytokine 27-plex Assay (Bio-Rad) according to the manufacturer’s protocol. Briefly, 50pl of serum samples were incubated with magnetic capture beads for 30 minutes, washed then incubated with detection antibody for 30 minutes, wash again, incubated with streptavidin- PE prior to assessment on the Bio-Plex 200 reader (Bio-Rad).

[0140] Assessment of IL-7 induction in murine ileal organoid in response to ionizing irradiation

[0141] Murine organoids were seeded in 6-well plates with Matrigel (Corning) on day 0. On day 5, these organoids were subjected to mock or a range dose (0.25-6Gy) of 320 kV X-ray at a dose rate of 0.87 Gy / minute using the XRad320 (X-ray Precision) irradiator. Culture medium was collected from the wells at 48 hours after IR. The concentration of murine IL-7 in these media was assessed by a commercial ELISA kit (Thermofisher) according to the manufacture’s guideline.

[0142] Combination of intestinal low dose irradiation (ILDR) and PD-L1 blockade in tumor models

[0143] All animal experiments were performed in compliance with French and European laws and regulations, and approved by the local institutional animal ethic board and French Ministere de la Recherche (permission number: 2023_047_43699). A total of 1 x 106MC38 cells in 100 pl PBS was inoculated subcutaneously into the right flank of 8-10 weeks old C57BL / 6J mice on day 0. Mice were subjected to intraperitoneal (i.p.) injection of isotype control (clone LTF-2, 10 pg / g of mouse body weight) or aPD-L1 Ab (clone 10F.9G2, 10 pg / g of mouse body weight) on day 6, 10, 14 and 18. Mice were also treated with mock or ILDR with the indicated dose on day 8. Mice were monitored for a minimum of 2 months since tumor inoculation for tumor growth kinetics, body weight and other adverse signs (e.g. bloody feces). In indicated experiment, fecal samples were collected from individual mice at day-1 before tumor inoculation (dayO), day4, day 7, day 11 and day18, whereas plasma was obtained on day 11 via submandibular sampling.

[0144] ILDR in the mouse model

[0145] ILDR in the murine tumor model was performed using the Small Animal Radiation Research Platform (SARRP, XStrahl). Mice were anesthetized with 2.5% of isoflurane on supine position. They were first subjected to a whole-body CT scan using cone beam computed tomography following by three-dimensional reconstruction of these CT images on the Treatment planning system (MuriPlan). An ILDR treatment plan was designed to deliver the prescribed RT dose (0.25Gy, 1Gy or 4Gy) to the intestinal zone while sparing the subcutaneous tumor on the flank. This was achieved by a 3-beam treatment plan at gantry angles of -135°, 70° and -10°, respectively with the isocenter placed on the center of the bowel area. The size of beam field was set at 14mm x 12mm with the moving variable collimator (MVC) encompassing the intestinal zone between liver and the bladder on coronal / sagittal plane. Tumors were spared from all beams with an approximate 5mm margin on the transverse plane. The out of field dose on the edge of these beams is < 10% as indicated by dosimetry calculation. ILDR was delivered using 220 KV X-ray at a dose rate of 3.021 Gy / min. Mice with mock irradiation were subjected to anesthesia for a similar period with the beam off.

[0146] Gut colonization of P. distasonis in C57BL / 6J mice

[0147] A strain of P. distasonis was isolated from Ileal mucosa tissue of a ORC patients via culturomics and stored in the bacterial biobank as previously described (Roberti et al, 2020). P. distasonis was grown on Columbia agar with 5% sheep blood plates in an anaerobic atmosphere created using anaerobic generators (Biomerieux) and harvested at exponential growth for their gut colonization in C57BL / 6J mice. In indicated experiments, mice were oral gavage with 100 pl of PBS or suspension containing 109P. distasonis bacteria (obtained using a fluorescence spectrophotometer at an optical density of 1 measured at a wave length of 600 nm) on day 5 and 7. The efficacy of colonization was confirmed by qPCR performed on fecal DNA with specific primers for P. distasonis 48 h after the first gavage.

[0148] Depletion of CD4+or CD8+T cells in mice

[0149] In indicated experiments, mice were subjected to i.p. injection of isotype contrl Ab (Rat lgG2b, clone LTF-2 50pg + Rat lgG2a, clone 2A3 50 pg), aCD4 Ab (clone GK1.5, 100 pg / mouse) or aCD8 Ab (clone 53-6.7, 100 pg / mouse) on day 5, 7, 9 and 13. The efficacy of T cell depletion was validated by flow cytometry assessment on blood samples obtained from these mice on day 11 via submandibular sampling. qPCR assessment of abundance of bacterial species in murine fecal samples

[0150] Fecal samples were obtained from mice in indicated experiments and subjected to DNA extraction using the QIAamp Fast DNA Stool Mini Kit (QIAGEN). The quality and quantity of DNA were measured by Nanodrop (Thermofisher). The abundance of indicated bacteria species in these samples was analyzed by qPCR using SYBR green and specific primer pairs (see Key resource table in Chen al, 2025). Reactions were run on ABI QuantStudio 5 Real-time PCR System using the standard condition (Thermo Fisher Scientific). Following the amplification, the threshold cycle values of targeted bacterial species and universal bacteria (reference for normalization) were recorded. The relative abundance of bacteria species was calculated using the ACt method.

[0151] RT-qPCR assessment of H7 in mouse ileal tissues

[0152] Ileal tissues were harvested from mice in indicated experiments and subjected to RNA extraction using the RNease mini kit with DNAse treatment (both from QIAGEN) to remove potential contamination of genomic DNA. The quality and quantity of mRNA were measured by Nanodrop. RNA samples were then reversed transcribed and analyzed by RT-qPCR using TaqMan probes specific for targeted genes (Thermofisher). The following genes were analyzed in this study: H7 and ppia (housekeeping gene).

[0153] Flow cytometry assessment

[0154] T umor and blood samples were harvested from mice. T umor were minced on the petri dish by scalpels and incubated in dissociation buffer (RPMI medium containing 2mg / mL collagenase IV, 0.2mg / mL Hyaluronidase and 150U / ml DNase I) at 37 °C for 45 minutes, then filtered through 100pM cell strainer, incubated with 1X RBC lysis buffer (BioLegend) on ice for 5 minutes, then run through the 40pM cell strainer. Following washing, cells were resuspended in FACS buffer (1X PBS containing 2% FBS, 1mM EDTA and 0.1% sodium azide) at a density of 2 x 106cells / 1 OOpl. These cells were first incubated with anti-mouse CD16 / 32 Ab (clone 93, Thermofisher) and Zombie Aqua Fixable Viability dye (BioLegend) dye on ice for 10 minutes, then a mixture of antibodies targeting surface markers. Following washing, cells were fixed and permeabilized using the Foxp3 I Transcription Factor Staining Buffer Set (Thermofisher) according to the manufacture’s protocol. They were then stained with antibodies recognizing intracellular markers. Anti-mouse fluorescence-conjugated Abs (clone) used for phenotyping were purchased from BioLegend, Thermofisehr, Miltenyi and Cell signaling Technology as follows: BV421 CXCR6 (SA051 D1), Vioblue CD4 (REA604), BV610 CD44 (IM7), BV650 CD25 (PC61), FITC CD8a (53-6.7), PE TCF-1 (C63D9), PE a4 7 (REA457), PE-Cy7 PD-1 (RMP1-30), PE-Cy7 RORyt (B2D), APC TOX (REA473), APC Foxp3 (FJK-16s), Alexa Fluor 700 CD45 (30-F11) and APC Vio770 CD3E (REA606). Upon washing, cells were resuspended in FACS buffer and analyzed by Cytoflex (Beckman). Flow cytometry data were analyzed by Flowjo (BD Bioscience) and the graphs were generated using Prism (GraphPad).

[0155] SABR in the spontaneous breast cancer model

[0156] Endogenous mammary tumors were established per conventional procedures (Buque et al, 2020). In brief, 6-7 weeks old female C57BL / 6J mice were implanted subcutaneously with 50 mg slow-release (90 days) medroxyprogesterone acetate (MPA, M) pellets (Innovative Research of America). Mice were then administered with 1 mg 7,12-dimethylbenz[a]anthracene (DMBA, D; Sigma-Aldrich) in 200 pL corn oil (Millipore Sigma) once a week on weeks 1 , 2, 3, 5, 6 and 7 after implantation of the MPA pellet. Mice were then routinely monitored for the emergence of neoplastic lesions along the mammary lines. Tumor surface was calculated as the area of an ellipse (A = longest diameter X shortest diameter X TT / 4), as per common procedures. Mice with primary tumor area reaching 15-45 mm2(dO) were randomized into subgroups to receive mock or SABR (20Gy x 2F or 20Gy x 3F). Following treatment, mice were monitored for tumor growth and the presence of secondary tumor. Mice were euthanatized when reached ethical endpoints (e.g. the cumulative tumor surface > 180-200 mm2, or in the presence of evident signs of toxicity or stress).

[0157] The DC-B3Z co-culture system

[0158] De-iniDCs were collected by trypsinization and diluted with DC medium containing p-mercaptoethanol and GM-CSF. Cells were seeded in 12-well plates (5x105cells / well) for a dose range (0-6Gy) of IR (intervention 1), or in 96-well p-bottom tissueculture plates for the treatment with a metabolite library of 96 compounds including bile acids / carnitines / sterols (at two concentrations i.e 2.5pg and 5pg / ml) for 16 hours (intervention 2), then pulsed with OVA protein (1mg / mL) for 4 hours and incubated with CTL (B3Z hybridoma cells) for 16 hours. The supernatant of cell culture was collected and the concentration of IL-2 in these supernatants was assessed by customized ELISA.

[0159] Table 1. The mean prevalence and relative abundance of significantly different species between Elite and non-Elite patients at V1. Abbreviations: Rs: responders; NRs: non-responders.

[0160] Table 2. The mean prevalence and relative abundance of contrasting species between OS 12 R and N R patients at V1. Abbreviations: Rs: responders; NRs: non-responders. Table 3. ANOSIM and PERMANOVA metrics and related P values from PCoA analysis of p diversity of microbiota in murine fecal samples collected at V2, V4 and V5.

[0161] Abbreviations: PCoA: principal component analysis; M: match; aPD-L1 : antiprogrammed death-ligand 1 ; Ab: antibody; ILDR: intestinal low dose irradiation; Gy: gray.

[0162] Table 4. Clinical characteristics of the first 10 patients enrolled in the ILDR perspective trial (NCT06076135).

[0163] REFERENCES

[0164] Beghini, F., McIver, L.J., Blanco-Miguez, A., Dubois, L., Asnicar, F., Maharjan, S., Mailyan, A., Manghi, P., Scholz, M., Thomas, A.M., et al. (2021). Integrating taxonomic, functional, and strain-level profiling of diverse microbial communities with bioBakery 3. Elife 10. 10.7554 / eLife.65088

[0165] Blanco-Miguez, A., Beghini, F., Cumbo, F., McIver, L.J., Thompson, K.N., Zolfo, M., Manghi, P., Dubois, L., Huang, K.D., Thomas, A.M., et al. (2023). Extending and improving metagenomic taxonomic profiling with uncharacterized species using MetaPhlAn 4. Nat Biotechnol 41, 1633-1644. 10.1038 / s41587-023-01688-w.

[0166] Buque, A., Bloy, N., Perez-Lanzon, M., Iribarren, K., Humeau, J., Pol, J.G., Levesque, S., Mondragon, L., Yamazaki, T., Sato, A., et al. (2020). Immunoprophylactic and immunotherapeutic control of hormone receptor-positive breast cancer. Nat Commun 11, 3819. 10.1038 / S41467-020- 17644-0.

[0167] Callahan, B.J., McMurdie, P.J., Rosen, M.J., Han, A.W., Johnson, A. J., and Holmes, S.P. (2016). DADA2: High-resolution sample inference from Illumina amplicon data. Nat Methods 13, 581-583. 10.1038 / nmeth.3869

[0168] Chen J, Levy A, Tian AL, Huang X, Cai G, Fidelle M, Rauber C, Ly P, Pizzato E, Sitterle L, Piccinno G, Liu P, Durand S, Mao M, et al. (2025) Low-dose irradiation of the gut improves the efficacy of PD-L1 blockade in metastatic cancer patients. Cancer Cell 43(3) :361-379.

[0169] Colbert, L.E., El Alam, M.B., Wang, R., Karpinets, T., Lo, D., Lynn, E.J., Harris, T.A., Elnaggar, J.H., Yoshida-Court, K., Tomasic, K., et al. (2023). Tumor-resident Lactobacillus iners confer chemoradiation resistance through lactate-induced metabolic rewiring. Cancer Cell 41, 1945-1962 e1911. 10.1016 / j.ccell.2023.09.012.

[0170] Derosa, L., Routy, B., Desilets, A., Daillere, R., Terrisse, S., Kroemer, G., and Zitvogel, L. (2021). Microbiota-Centered Interventions: The Next Breakthrough in ImmunoOncology? Cancer Discov 11, 2396-2412. 10.1158 / 2159-8290.CD-21-0236.

[0171] Durand, S., Grajeda-Iglesias, C., Aprahamian, F., Nirmalathasan, N., Kepp, O., and Kroemer, G. (2021). The intracellular metabolome of starving cells. Methods Cell Biol 164, 137-156. 10.1016 / bs.mcb.2021.04.001.

[0172] Fernandes, A., Oliveira, A., Soares, R., and Barata, P. (2021). The Effects of Ionizing Radiation on Gut Microbiota, a Systematic Review. Nutrients 13. 10.3390 / nu13093025.

[0173] Fernandes, A., Oliveira, A., Soares, R., and Barata, P. (2023). The Effects of Ionizing Radiation on Gut Microbiota: What Can Animal Models Tell Us?-A Systematic Review. Curr Issues Mol Biol 45, 3877-3910. 10.3390 / cimb45050249.

[0174] Fidelle, M., Rauber, C., Alves Costa Silva, C., Tian, A.L., Lahmar, I., de La Varende, A.M., Zhao, L., Thelemaque, C., Lebhar, I., Messaoudene, M., et al. (2023). A microbiota-modulated checkpoint directs immunosuppressive intestinal T cells into cancers. Science 380, eabo2296. 10.1126 / science.abo2296.

[0175] Fluckiger, A., Daillere, R., Sassi, M., Sixt, B.S., Liu, P., Loos, F., Richard, C., Rabu, C., Alou, M.T., Goubet, A.G., et al. (2020). Cross-reactivity between tumor MHC class I- restricted antigens and an enterococcal bacteriophage. Science 369, 936-942. 10.1126 / science.aax0701 .

[0176] Gopalakrishnan, V., Spencer, C.N., Nezi, L., Reuben, A., Andrews, M.C., Karpinets, T.V., Prieto, P.A., Vicente, D., Hoffman, K., Wei, S.C., et al. (2018). Gut microbiome modulates response to anti-PD-1 immunotherapy in melanoma patients. Science 359, 97-103. 10.1126 / science.aan4236.

[0177] Grajeda-Iglesias, C., Durand, S., Daillere, R., Iribarren, K., Lemaitre, F., Derosa, L., Aprahamian, F., Bossut, N., Nirmalathasan, N., Madeo, F., et al. (2021). Oral administration of Akkermansia muciniphila elevates systemic antiaging and anticancer metabolites. Aging (Albany NY) 13, 6375-6405. 10.18632 / aging.202739.

[0178] Guo, H., Chou, W.C., Lai, Y., Liang, K., Tam, J.W., Brickey, W.J., Chen, L., Montgomery, N.D., Li, X., Bohannon, L.M., et al. (2020). Multi-omics analyses of radiation survivors identify radioprotective microbes and metabolites. Science 370. 10.1126 / science.aay9097.

[0179] Hauer-Jensen, M., Denham, J.W., and Andreyev, H.J. (2014). Radiation enteropathypathogenesis, treatment and prevention. Nat Rev Gastroenterol Hepatol 11, 470-479. 10.1038 / nrgastro.2014.46.

[0180] Jadon, R., Higgins, E., Hanna, L., Evans, M., Coles, B., and Staffurth, J. (2019). A systematic review of dose-volume predictors and constraints for late bowel toxicity following pelvic radiotherapy. Radiat Oncol 14, 57. 10.1186 / s13014-019-1262-8.

[0181] Liu, X., Zhou, Y., Wang, S., Guan, H., Hu, S., Huang, R., and Zhou, P. (2019). Impact of Low-Dose Ionizing Radiation on the Composition of the Gut Microbiota of Mice. Toxicol Sci 171, 258-268. 10.1093 / toxsci / kfz144.

[0182] Ma, S., Shungin, D., Mallick, H., Schirmer, M., Nguyen, L.H., Kolde, R., Franzosa, E., Vlamakis, H., Xavier, R., and Huttenhower, C. (2022). Population structure discovery in meta-analyzed microbial communities and inflammatory bowel disease using MMUPHin. Genome Biol 23, 208. 10.1186 / s 13059-022-02753-4.

[0183] Minn, A. J., and Wherry, E.J. (2016). Combination Cancer Therapies with Immune Checkpoint Blockade: Convergence on Interferon Signaling. Cell 165, 272-275. 10.1016 / j. cell.2016.03.031.

[0184] Morad, G., Helmink, B.A., Sharma, P., and Wargo, J.A. (2022). Hallmarks of response, resistance, and toxicity to immune checkpoint blockade. Cell 185, 576. 10.1016 / j. cell.2022.01.008.

[0185] Pasolli, E., Schiffer, L., Manghi, P., Renson, A., Obenchain, V., Truong, D.T., Beghini, F., Malik, F., Ramos, M., Dowd, J.B., et al. (2017). Accessible, curated metagenomic data through ExperimentHub. Nat Methods 14, 1023-1024. 10.1038 / nmeth.4468.

[0186] Pitt, J.M., Vetizou, M., Daillere, R., Roberti, M.P., Yamazaki, T., Routy, B., Lepage, P., Boneca, I.G., Chamaillard, M., Kroemer, G., and Zitvogel, L. (2016). Resistance Mechanisms to Immune-Checkpoint Blockade in Cancer: Tumor-Intrinsic and -Extrinsic Factors. Immunity 44, 1255-1269. 10.1016 / j. immuni.2016.06.001. Ribas, A., and Wolchok, J.D. (2018). Cancer immunotherapy using checkpoint blockade. Science 359, 1350-1355. 10.1126 / science.aar4060.

[0187] Richter, C., Thieme, S., Bandola, J., Laugsch, M., Anastassiadis, K., and Brenner, S. (2013). Generation of inducible immortalized dendritic cells with proper immune function in vitro and in vivo. PLoS One 8, e62621. 10.1371 / journal. pone.0062621

[0188] Roberti, M.P., Yonekura, S., Duong, C.P.M., Picard, M., Ferrere, G., Tidjani Alou, M., Rauber, C., lebba, V., Lehmann, C.H.K., Amon, L., et al. (2020). Chemotherapy-induced ileal crypt apoptosis and the ileal microbiome shape immunosurveillance and prognosis of proximal colon cancer. Nat Med 26, 919-931. 10.1038 / s41591-020-0882-8.

[0189] Routy, B., Le Chatelier, E., Derosa, L., Duong, C.P.M., Alou, M.T., Daillere, R., Fluckiger, A., Messaoudene, M., Rauber, C., Roberti, M.P., et al. (2018). Gut microbiome influences efficacy of PD-1-based immunotherapy against epithelial tumors. Science 359, 91-97. 10.1126 / science.aan3706.

[0190] Ruaud, A., Esquivel-Elizondo, S., de la Cuesta-Zuluaga, J., Waters, J.L., Angenent, L.T., Youngblut, N.D., and Ley, R.E. (2020). Syntrophy via Interspecies H(2) Transfer between Christensenella and Methanobrevibacter Underlies Their Global Cooccurrence in the Human Gut. mBio 11. 10.1128 / mBio.03235-19.

[0191] Teng, H., Wang, Y., Sui, X., Fan, J., Li, S., Lei, X., Shi, C., Sun, W., Song, M., Wang, H., et al. (2023). Gut microbiota-mediated nucleotide synthesis attenuates the response to neoadjuvant chemoradiotherapy in rectal cancer. Cancer Cell 41, 124-138 e126. 10.1016 / j.ccell.2022.11.013.

[0192] Vetizou, M., Pitt, J.M., Daillere, R., Lepage, P., Waldschmitt, N., Flament, C., Rusakiewicz, S., Routy, B., Roberti, M.P., Duong, C.P., et al. (2015). Anticancer immunotherapy by CTLA-4 blockade relies on the gut microbiota. Science 350, 1079- 1084. 10.1126 / science.aad1329.

[0193] Zhao, L., Liu, P., Xie, W., Zhang, S., Thieme, S., Zitvogel, L., Kroemer, G., and Kepp, O. (2021). A genotype-phenotype screening system using conditionally immortalized immature dendritic cells. STAR Protoc 2, 100732. 10.1016 / j.xpro.2021.100732.

[0194] Zhao, L., Liu, P., Mao, M., Zhang, S., Bigenwald, C., Dutertre, C.A., Lehmann, C.H.K., Pan, H., Paulhan, N., Amon, L., et al. (2023). BCL2 inhibition reveals a dendritic cellspecific immune checkpoint that controls tumor immunosurveillance. Cancer Discov. 10.1158 / 2159-8290. CD-22- 1338.

[0195] Zitvogel, L., Ma, Y., Raoult, D., Kroemer, G., and Gajewski, T.F. (2018). The microbiome in cancer immunotherapy: Diagnostic tools and therapeutic strategies. Science 359, 1366-1370. 10.1126 / science.aar6918. B25-1157QT

[0196] 1 / 1

[0197] PCT

[0198] (Original in Electronic Form) (This sheet is not part of and does not count as a sheet of the international application)

[0199] FOR RECEIVING OFFICE USE ONLY

[0200] FOR INTERNATIONAL BUREAU USE ONLY

[0201] 0-5 This form was rec international Bure

[0202] 0-5- Authorized officer

Claims

CLAIMS1. A composition comprising live bacteria belonging to the Christensenellaceae family, for use in the treatment of cancer in a patient, in combination with intestinal low dose radiotherapy (ILDR) and an antineoplastic treatment.

2. The composition for use of claim 1 , wherein the composition comprises live bacteria of the genus Christensenella, preferably live bacteria of at least one of the species Christensenella minuta, Christensenella massiliensis, Christensenella timonensis, Christensenella hongkongensis, Christensenella intestinihominis and Christensenellaceae bacterium SGB14127 and more preferably live bacteria of the species Christensenella minuta, such as Christensenella minuta 601 deposited at the Collection Nationale de Cultures de Microorganismes (CNCM, Institut Pasteur, Paris) under the reference 1-6155.

3. The composition for use of claim 1 or claim 2, wherein the composition further comprises an archaea such as Methanobrevibacter smithii and / or Ruminococcus bromii bacteria.

4. The composition for use of any one of claims 1 to 3, wherein the composition is a live biotherapeutic product (LBP) comprising at least one cultured bacterial strain.

5. The composition for use of any one of claims 1 to 4, wherein the composition comprises comprises fecal microbiota material and is formulated to perform fecal microbiota transplant (FMT).

6. The composition for use of any one of claims 1 to 5, wherein said antineoplastic treatment comprises l-O therapy such as immune checkpoint blockade (ICB), preferably anti-PD-1 antibodies (Ab), anti-PD-L1 Ab, anti-PD-L2 Ab and / or anti- CTLA4 Ab.

7. The composition for use of any one of claims 1 to 5, wherein said antineoplastic treatment comprises chemotherapy selected from oxaliplatinum based- cytotoxicants, immunogenic oxaliplatinum based therapy, anthracyclins, FOLFIRINOX (5 Fluoro-uracile Irinotecan Oxaliplatine, 5 Fluoro-uracile Campto Eloxatine), taxanes and any other protocol for colon, pancreas, cervix, anal or pelvic carcinoma.

8. The composition for use of any one of claims 1 to 7, wherein said antineoplastic treatment comprises l-O therapy as recited in claim 6, combined with chemotherapy as recited in claim 7.

9. The composition for use of any one of claims 1 to 8, wherein said ILDR consists in exposing the patient’s intestines to at least 0.25 Gy and less than 4 Gy, preferably 0.5 to 3 Gy, more preferably about 1 Gy, in one fraction.

10. The composition for use of any one of claims 1 to 9, wherein said ILDR is administered to a volume of at least 7, 15, 50, 70, 90 or 100 cc and / or at most 1200, 900, 700, 600, 550, 500, 450 or 420 cc.11 . The composition for use of any one of claims 1 to 10, wherein said ILDR provokes an abscopal effect and / or favors the growth of immunogenic species such as Christensenella minuta and / or increases the patient’s response to said antineoplastic treatment.

12. A theranostic method for determining if a cancer patient needs a bacterial compensation before administration of a treatment combining ILDR and immunotherapy, comprising assessing, in a feces sample from said patient, the presence of bacteria belonging to the Christensenellaceae family, the Christensenella genus and / or the Christensenella minuta species, wherein if no bacteria belonging to the Christensenellaceae family the Christensenella genus and / or the Christensenella minuta species is present / detectable in said feces sample, the patient needs a bacterial compensation with the composition of any one of claims 1 to 5.

13. A method of treating cancer in a patient in need thereof, comprising administering an effective amount of the composition of any one of claims 1 to 5, in combination with ILDR and an antineoplastic treatment.

14. The method of claim 13, for treating a patient identified as in need of a bacterial compensation by a method of any one of claim 12.

15. A method of treating cancer in a patient in need thereof, comprising the following steps:(i) optionally, assessing the presence of (a) bacteria belonging to the Christensenellaceae family, (b) bacteria belonging to the Christensenella genus or (c) Christensenella minuta in a feces sample from said patient;(ii) optionally, especially if step (i) has been performed and has showed that the searched bacteria are not present, administering an effective amount of a composition as recited in any one of claims 1 to 5;(iii) performing ILDR as recited in claim 9 or claim 10;(iv) administering an antineoplastic treatment as recited in any one of claims 6 to 8; wherein step (iv) is performed between one day to three weeks, preferably less that 5 days after step (iii).

Images