Patent Application
20250020650
The present invention relates to a method of predicting a patient's benefit from therapy with an immune checkpoint inhibitor, a method of predicting a cancer patient's probability of survival, and a method for determining in a sample the value of an expression level of a platelet surface protein.
The present invention relates to the field of molecular biology and molecular medicine, more particular to the field of molecular diagnostics and prognostics.
Immune checkpoint receptors like CTLA4 and PD-1 are crucial for preventing excessive immune responses and autoimmunity. Seminal discoveries made by Allison and Honjo provided preclinical proof of concept data that blockage of CTLA4 and PD-1 signaling unleashes marked anti-tumor immune responses. Clinical evaluation revealed remarkable therapeutic potential of immune checkpoint inhibition in human cancer patients and for the first time allowed for long term survival of patients with advanced metastasized solid tumors. Besides melanoma patients, especially patients suffering from non-small cell lung cancer (NSCLC) benefit from treatment with antibodies inhibiting the PD-1 and CTLA4 immune checkpoints. Nevertheless, simple and robust biomarkers to predict therapy responses towards immune checkpoint inhibitor (ICI) are still missing.
With 1.8 million deaths per year, lung cancer represents one of the most frequent and lethal cancers worldwide. In the US 254, 170 new lung cancer cases are expected to be diagnosed in 2021. Given the high frequency of lung cancer and the cost of checkpoint inhibitory therapies, the lack of robust biomarkers to select patients who best possibly benefit from ICI represents a major burden for the health systems.
Histological quantification of intratumoral PD-L1 expression is routinely performed in an attempt to predict therapy responses towards ICI, however, only an insufficient correlation between detection of PD-L1 expression in tumor biopsies and the overall response rate (ORR) was found (see Teng F., Meng X., Kong L., Yu J. Progress and challenges of predictive biomarkers of anti PD-1/PD-L1 immunotherapy: A systematic review. Cancer letters. 2018; 414:166-173).
In lung cancer, evaluation of smoking history, tumor mutational burden (TMB), microsatellite instability (MSI), high expression of CTLA4, low expression of CX3CL1 and infiltration of CD8+ T cells within the tumor microenvironment (TME) seems to be superior in predicting therapy responses towards anti-PD-1/PD-L1 directed ICI when compared to histopathological PD-L1 quantification, however these markers so far could not be translated into a robust and clinically easy to use biomarker signature.
Interestingly, recent research results have shown that tumor-specific protein markers are not only present freely in the blood, but can also be taken up by platelets, so-called thrombocytes, which classically mediate hemostasis. Platelets express these protein markers, so that after their isolation an assessment on the course of tumor diseases can be made.
However, a distinctive feature of platelet surface protein expression is that a large number of them are regulated depending on the activation state of the platelet. For example, some proteins are stored preformed in intracellular vesicles and are transported to the cell surface upon platelet activation. Accordingly, the amount of protein on the surface increases in the activated state. In contrast, so-called matrix metalloproteases (MMPs) in activated platelets can cut proteins from the surface (so-called shedding), which reduces the total amount of protein on the platelet surface after activation.
However, since platelets exhibit different activation levels both during sample collection and in the organism due to various stimuli (negatively charged surfaces in the sample vessel, shear forces during blood collection, inflammatory mediators, etc.) and protein expression depends on the activation state, an exact determination of total protein expression on the platelet surface is difficult. Thus, platelet activation status acts as a confounding factor that reduces the sensitivity and specificity of platelet surface protein expression as a biomarker.
Against this background it is an object underlying the invention to provide a method for predicting a patient's benefit from therapy with an ICI, which involves the use of an appropriate biomarker. It is also an object underlying the invention to provide a method for predicting a cancer patient's probability of survival, where the prediction is based on an appropriate biomarker. By means of said methods the disadvantages of the state of the art are avoided or at least significantly reduced.
The present invention satisfies these and other needs.
The present invention provides a method of predicting a patient's benefit from therapy with an immune checkpoint inhibitor (ICI), comprising the steps of:
According to the invention, an “immune checkpoint inhibitor” (ICI) refers to a molecule that inhibits an immune checkpoint, i.e. key regulators of the immune system that when stimulated can dampen the immune response to an immunologic stimulus. ICIs can block inhibitory checkpoints, restoring immune system function. Currently approved checkpoint inhibitors target the molecules CTLA4, PD-1, and PD-L1. Typical immune checkpoint inhibitors are e.g. antibodies against CTLA-4 (e.g. ipilimumab), PD-1 (e.g. nivolumab) and PD-L1 (e.g. atezolizumab, durvalumab and avelumab). After the ICI has been infused, the ICI binds to these proteins, which act as immune checkpoints. As a result, the cells that carry one of these proteins on the cell surface and bind the ICI are temporarily (or as long as the therapeutic ICI is circulating in the body) attacked by immune cells and removed from the body by macrophages (temporary cell depletion). These processes lead to an intensification of the immune response against the tumor, so that its strategy of immune evasion is counteracted.
According to the invention a “platelets-containing sample” refers to any solution, liquid or semi-liquid sample, which contain thrombocytes or blood platelets, respectively, such as a blood sample, a buffered platelets solution, a solution of biological cells etc. The term platelet and thrombocyte can be used interchangeably.
According to the invention the expression level is a measure of the strength of the presence of the analyzed protein molecule in the platelets, i.e. the occurrence of the amount of protein exposed on the platelets' surface. In an embodiment of the invention the level of expression is indicated as the percentage of platelets which carry the analyzed protein molecule on their surfaces in proportion to all analyzed platelets. In a preferred embodiment of the invention the expression level of the protein molecule is determined via flow cytometry or fluorescence-activated cell sorting (FACS), respectively.
According to the invention “pPD-L1expr.” refers to the expression value or level of pPD-L1 as determined in step 2) of the method according to the invention.
According to the invention, “platelet programmed cell death 1 ligand 1” (pPD-L1), also known as cluster of differentiation 274 (CD274) or B7 homolog 1 (B7-H1), is a protein that in humans is encoded by the CD274 gene (Human=Entrez 29126; Ensembl: ENSG00000120217; UniProt: Q9NZQ7). Programmed death-ligand 1 (PD-L1) is a 40 kDa type 1 transmembrane protein that has been speculated to play a major role in suppressing the adaptive arm of immune systems during particular events such as pregnancy, tissue allografts, autoimmune disease and other disease states such as hepatitis. The binding of PD-L1 to the inhibitory checkpoint molecule PD-1 transmits an inhibitory signal based on interaction with phosphatases (SHP-1 or SHP-2) via immunoreceptor tyrosine-based switch motif (ITSM). This reduces the proliferation of antigen-specific T-cells in lymph nodes, while simultaneously reducing apoptosis in regulatory T cells (anti-inflammatory, suppressive T cells)—further mediated by a lower regulation of the gene Bcl-2.
According to the invention, a “platelet activation marker” or “A” refers to an indicator molecule which signals the activated status of the platelets. Platelet or thrombocyte activation is a complex morphological and biochemical change of platelets at the site of a vascular lesion, that initiates hemostasis. It is triggered primarily by the contact of platelet glycoproteins with subendothelial collagen fibers and von Willebrand factor. Examples of A include CD62P (P-selectin) and PAC-1 (first procaspase activating compound).
In the first sub-step of step 3) of the method according to the invention a matrix is provided having the following format:
Each row i is assigned to an expression level or value of A, e.g. the most upper row represents a high expression level of A, and, e.g., the most lower row represents a low expression level or value of A. Each column j is assigned to an expression level of pPD-L1, e.g. the most left column represents a low expression level or value of pPD-L1, and, e.g., the most right column represents a high expression level or value of pPD-L1.
In the second sub-step of step 3) of the method according to the invention the element aij is selected from the matrix based on the expression levels determined in step 2), i.e. Aexpr. and pPD-L1expr.. To do this, select row i, which indicates the value or range of values of A corresponding to Aexpr., and select column j, which indicates the value or range of values of pPD-L1 corresponding to pPD-L1expr.. Then determine from the pair of values the corresponding pPD-L1corr. from the matrix.
According to the invention, in step 4) the values of pPD-L1expr. and pPD-L1corr. are added to result in pPD-L1adj..
In step 5) of the method according to the invention the prediction of the responsiveness of the patient to the ICI therapy is made. The inventors have found out that there will be a therapeutic benefit if pPD-L1adj. is equal or superior to a reference value x, and there will be less or even no therapeutic benefit if pPD-L1adj. is lower than a reference value x. The reference value x can be empirically determined by the skilled person by taking the median of pPD-L1adj. in the entire reference cohort.
The object underlying the invention is herewith fully achieved.
The inventors have realized that by means of the provided method an exact determination of a platelet expressed biomarker, pPD-L1, can be realized, independent of the activation state of the platelets. This is made via determining a corrected expression level pPD-L1corr. of pPD-L1. By adding pPD-L1corr. to the actual measured value of pPD1 the adjusted expression level pPD-L1adj. is obtained. Once this latter value is equal to or exceeds a threshold or reference value x, the patient has a positive prognosis for a ICI therapy.
As a consequence, the method according to the invention helps to identify the kinds of patients which are very likely to have a good responsiveness to an ICI therapy and, therefore, those patients which benefit best from such a therapy. On the other hand, the method allows excluding those patients from an ICI therapy which will not have significant benefit from an ICI therapy. It helps to reduce the burden on the healthcare system, as ICI therapy is very expensive. It further prevents patients with a low responsiveness from the side effects which are associated with an ICI therapy. In such a situation the side effects are not in a tolerable relationship with the expected therapy success.
In an embodiment of the method according to the invention, A is CD62P and, consequently, Aexpr. is CD62Pexpr..
This measure has the advantage that a particularly suited platelet activation marker is used which finally results in a reliable prognosis of the patient's responsiveness.
CD62P, also called P-selectin, granule membrane protein 140 (GMP-140), and platelet activation-dependent granule to external membrane protein (PADGEM), is a type-1 transmembrane protein that in humans is encoded by the SELP gene (Entrez: 6403; Ensembl: ENSG00000174175; UniProt: P16109). CD62P functions as a cell adhesion molecule (CAM) on the surfaces of activated endothelial cells, which line the inner surface of blood vessels, and activated platelets. In inactivated platelets CD62P is stored in α-granules.
In another embodiment of the invention in the matrix n is 4 resulting in pPD-L1 quartile groups Q1 (very low), Q2 (low), Q3 (high), and Q4 (very high).
The inventors recognized that subdividing in the matrix the expression levels of pPD-L1 into quartiles is useful and allows a sufficiently accurate prediction of the patient's response rate to ICI therapy.
In yet another embodiment of the method according to the invention in the matrix
According to the inventors' findings, these subdivisions have proven to be expedient and are therefore particularly advantageous.
In another embodiment of the invention in the matrix m is 5 resulting in CD62P expression groups E1, E2, E3, E4, and E5.
It turned out that the subdivisions of the expression levels of CD62P into 5 groups is adequate and results in a proper process of determining of pPD-L1corr..
In still another embodiment of the method according to the invention in the matrix
The inventors realized that the allocation of the indicated expression ranges is reasonable for an accurate determination of pPD-L1corr..
In another embodiment of the invention the elements aij of the matrix are as follows:
As a result, the matrix according to the invention looks as follows:
The values of the respective aij were empirically determined by the inventors and turned out to allow an accurate determination of the correction value pPD-L1corr..
In a preferred embodiment of the invention the reference value x is 2.1%.
The reference value of x=2.1% has been identified by the inventors as being of particular preference because it allows a reliable determination of the responsiveness of the patients to ICI therapy.
In another embodiment of the invention, the expression level of pPD-L1expr. and/or Aexpr. is determined via flow cytometry, preferably via fluorescence-activated cell sorting (FACS).
The use of flow cytometry or FACS allows an accurate and a time saving determination of the expression levels of pPD-L1expr. and Aexpr.. These techniques enable an implementation of the method according to the invention in the everyday routine of clinical laboratories.
In another embodiment of the invention said platelets-containing sample is a blood sample.
This measure has the advantage that the optimum sample of platelets is used, which routinely occurs anyway with the patients being treated. Therefore, performing the method according to the invention does not impose any additional burden on the patient.
In yet another embodiment of the invention said immune ICI is selected from the group consisting of: pembrolizumab, nivolumab, ipilimumab, tremelimumab, cemiplimab, spartalizumab, atezolizumab, durvalumab, and avelumab.
By this measure the measure according to the invention is adapted to the ICIs which are currently available. This embodiment, therefore, implements the method into the common ICI therapies.
In a still further embodiment of the invention said patient is suffering from non-small cell lung cancer (NSCLC).
This measure has the advantage that the method according to the invention is applied to a cancer disease which is currently treated with ICIs, and in which the method according to the invention has been shown to allow a particularly reliable prediction of therapeutic success.
Another subject-matter of the invention relates to a method for predicting a cancer patient's probability of survival, comprising the steps of:
The features, characteristics, advantages and embodiments of the therapy-responsiveness prediction method according to the invention mentioned at the outset apply—mutatis mutandis—also to the present cancer-progression prediction method. In contrast to the therapy prediction method in step 5) a poor survival prognosis is diagnosed if pPD-L1adj. is equal or superior to the reference value x, and a good survival prognosis is diagnosed if pPD-L1adj. is lower than the reference value x.
Still another subject-matter of the present invention relates to a method for determining in a sample the value of an expression level of a platelet surface protein, said expression level being independent of the activation state of the platelet (pind.), said method comprising the following steps:
The features, characteristics, advantages and embodiments of the therapy-responsiveness prediction method according to the invention mentioned at the outset apply—mutatis mutandis—also to the present expression determining method.
The inventors have realized that the principles embodied in the therapy-responsiveness prediction method with respect to pPD-L1 can be likewise applied to other platelet surface proteins to determine the “real” expression level of said platelet surface proteins.
It is to be understood that the before-mentioned features and those to be mentioned in the following cannot only be used in the combination indicated in the respective case, but also in other combinations or in an isolated manner without departing from the scope of the invention.
The invention is now further explained by means of embodiments resulting in additional features, characteristics and advantages of the invention. The embodiments are of pure illustrative nature and do not limit the scope or range of the invention. The features mentioned in the specific embodiments are general features of the invention which are not only applicable in the specific embodiment but also in an isolated manner and in the context of any embodiment of the invention.
The invention is now described and explained in further detail by referring to the following non-limiting examples and figures.
During 2016-2019, 173 consecutive patients with non-small lung cancer (NSCLC) treated in the Department of Medical Oncology and Hematology and Department of Internal Medicine VIII, University Hospital Tuebingen, Germany were prospectively included in the study (screening cohort=SC). In order to preclude the influence of anticoagulants like aspirin (ASS), low molecular weight heparin (LMWH) or other heparinoids and non-vitamin K antagonist oral anticoagulants (NOACs), long-term medication of each patient was considered. In the inventors' cohort 12 patients with LMWH and 28 patients taking ASS and/or clopidogrel were excluded. In
Written informed consent was given in all cases. This study was approved by IRB (ethics committee of the Faculty of Medicine of the Eberhard Karls University Tuebingen) and of the University Hospital Tuebingen and was conducted in accordance with the Declaration of Helsinki; reference number 456/BO2.
Platelets were obtained from healthy donors (not taking any medication for at least 10 days) and NSCLC patients after informed writing consent. Citrated blood was briefly centrifuged for 20 min at 120×g, the upper fraction was harvested as platelet-rich plasma (PRP). Platelets were washed twice with citrate wash buffer (128 mmol/L NaCl, 11 mmol/L glucose, 7.5 mmol/L Na2HPO4, 4.8 mmol/L sodium citrate, 4.3 mmol/L NaH2PO4, 2.4 citric acid, 0.35% bovine serum albumin, and 50 ng/ml prostaglandin E1 (PGE1)). To avoid the influence of PGE1 on platelet-tumor cell and platelet-immune cell interaction, the inventors did not use PGE1 in their co-incubation experiments. For platelet activation 10 μM of the Thrombin Receptor Activator Peptide 6 (TRAP-6), a protease-activated receptor 1 (PAR1) agonist, 2.5 μM ADP or 5 μg/mL Collagen was added to the platelets for 2 minutes. Platelets were fixed by 2% paraformaldehyde for 10 min and washed twice with PBS containing 1% FCS.
Flow cytometry was performed using fluorescence-conjugates or specific mAb and their controls followed by species-specific conjugate (Table 2) using a FACS Cantoll flow cytometer (Beckman Coulter) or a LSRFortessa (Becton Dickinson) from the flow cytometer facility Tuebingen.
Positive cells in percentage (%) were calculated as follows: Surface expression in percent obtained with the specific antibody—surface expression in percent obtained with isotype control. Platelets were preselected by CD41a+ and CD62P− (resting) or CD62P+ (activated). Data analysis was performed using FlowJo software (v.10). In order to verify the reproducibility of the inventors' flow cytometry system, the inventors performed a Bland-Altman analysis (
Tissue samples were fixed in 4% formalin and paraffin-embedded (FFPE) at the Department of Pathology (University Hospital Tuebingen). The sections were cut briefly in 3 μm sections and stained with Hematoxylin/Eosin and CD61 (clone: 2C9.G3) following standard protocols. For immunofluorescence microscopy, sections were deparaffinized and hydrated in a first step. The heat-induced antigen retrieval method was performed using sodium citrate buffer (pH 6.0) for 30 min. Antigen blocking was performed with Blocking solution (Zytomed) for 60 min. Primary antibodies that were included anti-CD41, mouse, 1:250 (clone: HIP8) and anti-PD-L1, rabbit, 1:200 (clone: 28-8). Secondary antibodies include Alexa-Fluor 594 labelled anti-rabbit (1:1000, Invitrogen) and Fluor 488 labelled anti-mouse (1:1000, Invitrogen). DAPI (1:1000, BioLegend) was used for nuclear staining prior mounting the slides with H-1500 Vectashield Hardset. Microscopic analysis was done with an Olympus BX63 microscope and a DP80 camera (Olympus).
For immunostaining tumor cells and/or platelets were fixed in 2% PFA in PBS (pH 7.4) for 10 min at −20° C. After three washing steps in PBS cells were incubated with a BSA blocking solution (2% BSA, 0.2% Triton X-100, 0.1% Tween) for 1 hour. Primary antibodies were anti-PD-L1, rabbit (1:250, clone: 28-8), anti-CD41, mouse (1:400, clone: HIP8), anti-CD61, rabbit (1:250, clone: SJ-19-09), anti-GFP, rabbit (1:200, clone: EPR14104), anti-fibronectin, mouse (1:200, clone: P1H11); as secondary antibodies Alexa-Fluor 488/594 labelled anti-rabbit (1:1000, Invitrogen) and Fluor 488/594 labelled anti-mouse (1:1000, Invitrogen) were used. Afterwards slides were mounted in fluorescent mounting medium containing DAPI (1:1000, BioLegend) counter-stain. For the plasma membrane staining CellMask™ (ThermoFisher) and DiI (ThermoFisher) was used according to manufactures instructions. For nuclear staining NucBlue™ (ThermoFisher) was used. Image acquisition was performed using an Olympus BX63 microscope and a DP80 camera (Olympus). Quantification of platelets, fibronectin and tumor cells were performed via counting fluorescence positive signals using an ImageJ script (v.1.52).
Paraffin-embedded patient samples were cut in 2-5 μm slices and collected on object slides. Subsequently, sections were subjected to deparaffinization and rehydration. Slides were treated with xylene for 10 mins, followed by rehydration using an ethanol dilution series of 100%, 95%, 70%, 50% for 5 mins each. One last change was performed using deionized water. Heat-induced antigen retrieval was performed using a Sodium-Citrate buffer (10 mM Sodium citrate, 0.05% Tween 20, pH 6.0) and boiling the samples for 20 mins. Samples were cooled down and stored in MACSima™ Running Buffer (Miltenyi Biotec, 130-121-565) until initial DAPI staining (Miltenyi Biotec, 130-111-570). The MACSima™ device is an ultra-high content cyclic IF device which allows for fully automated IF imaging. Iteratively, the device performs fluorescent staining with multiple labelled antibodies, image acquisition, and bleaching per cycle. Images were generated according to the manufacturer's instructions and analyzed with the Qi Tissue Image Analysis Software. For quantification at least two ROI were selected based on manual prestaining of DAPI.
For live cell imaging analysis A549 cells (cultured as stated above) were used. Tumor cells were co-incubated with platelets at a platelet-tumor cell ratio of 1:1000. Platelets were added to the tumor cells directly prior image acquisition. Platelet-tumor cell interaction was analyzed using confocal microscopy at frame intervals of 30 seconds for up to 40 min (Leica SP8 SMD; Leica HC PL APO CS 40×/0.85) using a fixed focus. Cell positions were assigned by their center-of-mass coordinates.
For transmission electron microscopy, platelets from one representative pPD-L1 high expressing NSCLC patient were used. Platelets were centrifuged and the resulting pellets were fixed for 24 hours in Karnovsky's fixative. As previously described, Ultrathin sections were examined with a LIBRA 120 (Zeiss) operating at 120 kV. For immunoelectron microscopy, platelets were fixed and embedded in Lowicryl K4M (Polysciences). Samples were stained with anti-PD-L1 antibody (Abcam) and examined using a LIBRA 120 transmission electron microscope (Zeiss) at 120 kV.
Protein levels of PD-L1 were measured using a human PD-L1 ELISA kit (abcam, clone: 28-8) according to the recommendations of the manufacturer. All concentrations are expressed as means±SEM of triplicates.
Whole-cell extracts were prepared using RIPA buffer and protein concentration was analyzed using the BioRad Dc assay. 25-50 μg of protein were transferred to 10-15% SDS-Page and blotted on a PVDF membrane (Millipore) with a wet blot system. The membrane was blocked for 1 h at room temperature with Roti-Block, followed by overnight incubation with the following antibodies: anti-PD-L1, rabbit (1:2000, clone: 28-8), anti-fibronectin, mouse (1:250, clone: P1H11), anti-Vinculin, mouse (1:10,000, clone: hVIN-1), anti-α tubulin (1:10,000, clone 11H10) and anti-β Actin (1:10,000, clone AC-15). Blots were visualized using ECL reagents (GE Healthcare) or the Super Signal West Kit (Thermo Scientific) and the ChemiDoc™ MP Imaging System.
To determine mRNA abundance in several tumor cell lines the inventors extracted mRNA in TriFast (Peqlab) according to the manufacturer's instructions. After DNAse digestion reverse transcription of total RNA was performed using random hexamers (Roche Diagnostics) and SuperScriptII reverse transcriptase (Invitrogen). Amplification of the respective genes by real-time polymerase chain reaction (RT-PCR) was performed in a total volume of 20 μl using 40 ng of cDNA, 500 nM forward and reverse primer and 2× GoTaq® qPCR Master Mix (Promega) according to the manufacturer's protocol. Cycling conditions were as follows: initial denaturation at 95° C. for 2 min, followed by 40 cycles of 95° C. for 15 sec, 55° C. for 15 sec and 68° C. for 20 sec. For amplification the following primers were used (5′->3 orientation): Fibronectin (FN1), fw ACCGTGGGCAACTCTGTCAA (SEQ ID NO: 1), rev CCCACTCATCTCCAACGGCA (SEQ ID NO: 2); Tissue factor (F3), fw GGCACGGGTCTTCTCCTACC (SEQ ID NO: 3), rev TGTCCGAGGTTTGTCTCCAGG (SEQ ID NO: 4); Von Willebrand Factor (VWF), fw CCTGCACCGACATGGAGGAT (SEQ ID NO: 5), rev CGTAAGTGAAGCCCGACCGA (SEQ ID NO: 6); Fibrinogen A (FBG), fw TGAAACGACTGGAGGTGGACA (SEQ ID NO: 7), rev CACGAGCTAAAGCCCTACTGC (SEQ ID NO: 8); GAPDH (GAPDH), fw TCGACAGTCAGCCGCATCTT (SEQ ID NO: 9), rev GCCCAATACGACCAAATCCGT (SEQ ID NO: 10). Real-time PCR amplifications were performed on a CFX96 Real-Time System (Biorad) and all experiments were performed in duplicate. The housekeeping gene GAPDH was used to standardize the amount of sample RNA.
Tumor cells were coated with platelets as described previously with slight modifications. Briefly summarized, PRP was obtained from fresh whole blood by centrifugation for 20 minutes at 120 g. Platelets were washed twice with citrate wash buffer (128 mmol/L NaCl, 11 mmol/L glucose, 7.5 mmol/L Na2HPO4, 4.8 mmol/L sodium citrate, 4.3 mmol/L NaH2PO4, 2.4 citric acid, 0.35% bovine serum albumin). In some experiments platelets were pretreated with 5 μg/mL anti-CD42b (clone: AK2), 20 μg/mL anti-integrin β1 (clone: P4C10) and anti-Integrin α5 (clone: JBS5) or corresponding control IgG1 (20 μg/mL) for 30 minutes at 37° C. and 7% CO2. Tumor cells were incubated in platelets at a platelet-tumor cell ratio of 1:1000 for 30 minutes at 37° C. and 7% CO2. For immunofluorescence microscopy and FACS analysis cells were fixed in 2% PFA in PBS (pH 7.4) for 10 min at −20° C. prior staining.
To prepare fibronectin matrices, plates were coated with a human plasma fibronectin purified protein (R&D, Minneapolis, MN, USA), (concentration 50 μg/cm2) for 120 min. For blocking of GPIb-IX-V complex, α5β1 or GPIIbIIIa, washed platelets (8×107/mL) were pretreated with 5 μg/mL anti-CD42b (clone: AK2), 20 μg/mL anti-integrin β1 (clone: P4C10) and anti-Integrin α5 (clone: JBS5), 1 μg/mL Tirofiban or corresponding control IgG1 (20 μg/mL) for 30 minutes at 37° C. and 7% CO2. After co-incubation with platelets (8×107/mL) for 30 min at 37° C. and 7% CO2, non-adherent platelets were removed via three washing steps using PBS. After removal of non-adherent platelets cells were fixed in 2% PFA in PBS (pH 7.4) for 10 min at −20° C. Platelet adhesion to fibronectin fibrils was evaluated by calculating surface coverage area and platelet count/FoV and from microscopic images using an ImageJ script (v. 1.52).
For overexpression of PD-L1 (CD274) a True-ORF-GFP-tagged expression vector was used (OriGene Technologies, Rockville, MD, USA). Control cells were transfected using a FLAG tag. The FLAG cDNA was generated by PCR and cloned into the PD-L1-GFP vector using AsiSI and MluI restriction sites. Tumor cells were transfected with 2.5 μg DNA (PD-L1-GFP, FLAG-GFP) using Lipofectamine™ 3000, in accordance to the manufacturer's instructions.
Peripheral blood mononuclear cells (PBMC) from healthy donors were isolated using Ficol/Paque (Biochrom) density gradient centrifugation after informed consent. All tumor cell lines were cultured with 10% FCS in Roswell Park Memorial Institute (RPMI) 1640 Medium at 37° C. and 7% CO2. Cell proliferation was quantified using a Neubauer chamber; for viability testing Trypan blue staining's was performed using a 0.4% trypan blue solution (Fluka). The tumor cell lines A549, NCI-H460, NCI-H23, NCI-H226, NCI-H322, NCI-H522, HOP-62 and HOP-92 were obtained from the American Type Culture Collection (ATCC). Mycoplasma contamination was excluded via a PCR-based method.
Freshly thawed (ex vivo) PBMCs from healthy donors were analyzed by enzyme-linked immunospot (ELISPOT) assay in duplicates. Interferon γ (IFNγ) ELISPOT assays in the inventors' study were performed as described previously. In brief, 96-well nitrocellulose plates (Millipore) were coated with 1 mg/mL anti-IFNγ mAb (Mabtech) and incubated overnight at 4° C. In a next step, plates were blocked with human serum (10%) for 2 hours at 37° C. PBMCs (2.5×105 cells per well) were pulsed with an EBV/CMV epitope mix containing the frequently recognized peptides BRLF109-117 YVLDHLIVV (SEQ ID NO: 11) (A*02) peptide and CMV pp65 (A*02) peptide NLVPMVATV (SEQ ID NO: 12) and incubated with or without platelets (ratio 1:50) for 24 hours. Phytohemagglutinin was used as positive control. HLA-A*02 (KLFEKVKEV (SEQ ID NO: 13))- and B*07 (KPSEKIQVL (SEQ ID NO: 14))-restricted control peptides derived from benign tissues (HV-exclusive HLA ligands) served as negative control. Prior co-incubation with T cells PD-L1 positive platelets from NSCLC patients were pre-treated with the anti-PD-L1 mAB Atezolizumab for 30 min and washed twice with PBS containing 1% FCS. Readout was performed according to the manufacturer's instructions. Spots were counted using an ImmunoSpot S5 analyzer (CTL).
Peptide-specific T cells were further analyzed by intracellular cytokine and cell surface marker staining. PBMCs were incubated with 10 μg ml−1 of peptide, 10 μg ml−1 brefeldin A (Sigma-Aldrich) and a 1:500 dilution of GolgiStop (BD) for 12-16 h. Staining included Cytofix/Cytoperm solution (BD), anti-CD4, mouse (1:100, clone: RPA-T4), anti-CD8, mouse (1:400, clone: B9.11), anti-TNF, mouse (1:120, clone: Mab11) and anti-IFN-γ, mouse (1:200 dilution, clone: 4SB3). PMA (5 μg ml−1) and ionomycin (1 μM, Sigma-Aldrich) served as positive control. Viable cells were determined using Aqua live/dead (1:400 dilution, Invitrogen). Samples were analyzed on a FACS Canto II cytometer (BD) and evaluated using FlowJo software v. 10.0.8 (BD).
The generation of NY-ESO-1-specific T cells was performed using as described previously. In brief, PBMCs from a healthy donor (1×107/mL) were stimulated using pools of NY-ESO-1 overlapping peptides (1 μg/mL). The NY-ESO-1 overlapping peptide pool of 15 amino acid length (11 amino acid overlap) was purchased via Miltenyi Biotec. The cells were cultured in RPMI 1640 containing 10% human AB-serum and 1% L-glutamin in the presence of 10 U/mL recombinant IL-2 and 10 ng/ml II-7. Culture medium was replaced every third day. After a pre-sensitization period of 7-14 days, NY-ESO-1 specific, IFNγ+ T cells were enriched after re-stimulation with NY-ESO-1 peptide pool for 6 h using CliniMACS® (Miltenyi Biotec) technique as reported previously. After enrichment, NY-ESO-1 specific T cells were expanded for 14 days in the presence of IL-7 (10 ng/ml), IL-15 (10 ng/mL) and IL-2 (50 U/mL). T cell specificity was analyzed via intracellular IFNγ staining as stated above. For further characterization of the T-cells the differentiation markers anti-CD45RO, mouse (1:200, clone: HI100), anti-CD62L, mouse (1:400, clone: DREG-56), anti-CD28, mouse (1:200, clone: CD28.2) and anti-CD27, mouse (1:200, clone: M-T271) were co-analyzed by flow cytometry. For the platelet-T-cell co-incubation assay, NY-ESO-1 specific T cells (5×106/mL) were cultured in TexMACS GMP Medium (Miltenyi Biotec). Six hours prior analysis T cells were co-incubated with platelets of NSCLC patients or healthy donors (ratio 1:200) and re-stimulated with NY-ESO-1 peptides (1 μg/mL). In order to investigate the functional role of PD-L1 on platelets surfaces, PD-L1 positive platelets from NSCLC patients were pre-treated with Atezolizumab (100 μg/mL) for 30 min and washed twice with PBS containing 1% FCS. As a negative control a Myelin oligodendrocyte glycoprotein (MOG) peptide mix was used (1 μg/mL). SEB (Toxin Technology, Sarasota, FL, USA) at 10 μg/mL was used as positive control. NY-ESO-1 specific T cell activity was determined by intracellular TNFα and IFNγ quantified via flow cytometry as described above.
HOP-62 and NCI-H23 cells were grown on glass bottomed plates. After two washing steps cells were fixed in 1% PFA in PBS (pH 7.4) for 10 min at −20° C. After three washing steps in PBS cells were incubated with a BSA blocking solution (5% BSA, 0.2% Triton X-100, 0.1% Tween) for 30 min. In situ PLA was performed using the Duolink PLA kit (Sigma-Aldrich) according to the manufacturer's instructions. In brief, after blocking cells were incubated with anti-PD-L1, rabbit (1:250, clone: 28-8) and anti-fibronectin, mouse (1:200, clone: P1H11) for 2 h at room temperature. After three washing steps with PBST (phosphate buffered saline, 0.1% Tween), anti-mouse PLUS and anti-rabbit MINUS PLA probes were linked to the primary antibodies for 1 h at 37° C. After three times washing steps with buffer A (0.01 M Tris, 0.15 M NaCl, and 0.05% Tween-20), PLA probes were ligated for 60 min at 37° C. After two washing steps with buffer A, amplification using Duolink In Situ Detection Reagents (Sigma) was performed at 37° C. for 120 min. Following amplification, cells were washed three times for 5 min with wash buffer B (0.2 M Tris 0.1 M NaCl). Cells were than coated with Duolink Mounting Medium containing DAPI. Image acquisition was performed using an Olympus BX63 microscope and a DP80 camera (Olympus).
Since platelet pre-activation levels differ due to sample collection/preparation and protein surface expression depends on the platelet activation state, accurate determination of total protein expression on platelet surfaces is challenging. As a result, the platelet pre-/activation level acts as a confounding factor and thus impairs the suitability of pPD-L1 as a promising biomarker in NSCLC. To circumvent this dilemma, the inventors established an activation-independent calculation matrix of platelet PD-L1. The matrix is based on the inventors' cohort of 128 NSCLC patients and investigates the activation-dependent expression change of PD-L1 (ΔpPD-L1) during controlled platelets stimulation ex vivo. Patients were categorized into pPD-L1 quartile groups (Q1: very low, Q2: low, Q3: medium, Q4: high), according to the pPD-L1 expression in unstimulated platelets. The pre-activation of platelets after sample preparation was determined via CD26P expression. In a second step each quartile group was subdivided according to the respective pre-activation levels (CD62P expression: 0-20%, 20-40%, 40-60%, 60-80%, and 80-100%) according to the pre-activation levels. The activation-dependent expression changes of PD-L1 (ΔpPD-L1) was then calculated for each subgroup. An overview of the subsampling and calculation is given in
Student's t test, Mann-Whitney U test, one-way ANOVA and Friedman's test were used for continuous variables, chi-squared test or Fisher's exact test for categorical variables. If significant differences by ANOVA were found, group wise comparison was done (Tukey's multiple comparison test). If significant differences by Friedman's test were found Dunn's multiple comparisons test was used. Overall survival (OS) and progression free survival (PFS), including the median, were calculated using the Kaplan-Meier method. Hazard ratios (HRs) were determined using Cox regression analysis. OS was calculated from the date of primary diagnosis or time-point of study inclusion and stratified by the end of the study. The predictive value of platelet-derived PD-L1 as a prognostic factor was evaluated by examining the area under the receiver-operator characteristic (ROC) curve using a confidence interval of 95%. All statistical tests were considered statistically significant when P was below 0.05. Statistical analysis was performed using SigmaStat, version 21 (SPSS) and GraphPadPrism (v.8.1.0).
To address whether the immune regulatory protein PD-L1 can be transferred from tumor cells to platelets, the inventors co-incubated platelets obtained from healthy donors with four different NSCLC tumor cell lines harboring varying expression levels of PD-L1 (NCI-H23, A549: PD-L1 low/negative, NCI-H226, NCI-H460: PD-L1 positive) (
Of note, conditioned medium from tumor cells induced platelet activation but did not result in increased levels of PD-L1 protein on the platelet surface (
To gain deeper insights into the interaction of platelets and lung cancer cells, the inventors took advantage of a live cell imaging platform, where platelets are added to the medium and circulate through an imaging chamber that contains human NSCLC cells. Real time video microscopy revealed distinct interactions of tumor cells and platelets (
While platelets are a nuclear, protein translation from RNA can nevertheless occur within platelets. The inventors therefore set out to investigate whether PD-L1 expression in platelets depends on a transfer of PD-L1 protein from tumor cells to platelets or whether a transfer of PD-L1 mRNA with subsequent protein synthesis within the platelet is involved. Transfection of vectors encoding for PD-L1-GFP and FLAG-GFP fusion proteins into PD-L1 negative A549 cells (
While the transfer of PD-L1-GFP or FLAG-GFP was robustly observed across various NSCLC cell lines, the inventors nevertheless noted differences in protein transfer efficacies. For example, platelets showed low levels of FLAG-GFP and PD-L1-GFP after co-incubation with NCI-H322, NCI-H522 and NCI-H23 cells, while HOP-62 and HOP-92 cells displayed significantly higher protein transfer rates (
To address the significance of the inventors' findings for human cancers, the inventors next quantified PD-L1 expression on platelets in healthy lung tissue or NSCLC tumor tissue. While platelets were detected in high abundance in healthy lung tissue and PD-L1 negative NSCLC, the inventors could not observe any relevant PD-L1 expression on these platelets (
Prompted by these results, the inventors next explored whether pPD-L1 exerts immune-inhibitory functions. The inventors stimulated human T cells from healthy donors with EBV/CMV-derived peptides in the presence or absence of PD-L1 positive platelets obtained from NSCLC patients. T cell activation was evaluated using an enzyme-linked-immuno-Spot (ELISpot) assays determining the effector cytokines IFNγ and TNFα. In line with published data the inventors observed that platelets dampen T cell activity independent of their PD-L1 expression status (
To investigate a potential impact of PD-L1 positive platelets on other immune cells, the inventors also characterized changes in the overall immune cell composition (peripheral blood) in 10 NSCLC patients and five healthy controls (
Regulation of pPD-L1 During Platelet Activation
As it is well established that expression levels of platelet surface proteins correlate with the platelet activation status, the inventors reasoned that different degrees of platelet activation might underlie varying levels of pPD-L1 expression on the platelet surface. Indeed, when the inventors analyzed the platelet activation marker CD62P, the inventors observed varying CD62P expression levels which showed a strong positive correlation with pPD-L1 expression (
As even highly standardized blood collection procedures can result in varying levels of shear-stress mediated platelet activation and therefore complicates standardization, the inventors hypothesized that different levels of platelet pre-activation might complicate the interpretability and comparability of pPD-L1 levels on freshly collected platelets from different patients. The inventors therefore reasoned that a controlled in vitro activation of platelets with subsequent maximization of pPD-L1 expression might most adequately uncover the total payload of platelet PD-L1 and best possibly allow a comparison between different patients. Indeed, the inventors found that pPD-L1 expression was maximized upon controlled platelet stimulation with the PAR1 agonist TRAP-6 (
The inventors first used the calculated pPD-L1adj. levels and performed a receiver operating characteristic (ROC) analysis for overall survival (OS). The inventors found that pPD-L1adj. levels in the subgroup of maximal platelet activation (CD62P 80-100%) showed highest accuracy in predicting OS (
It has been reported that mutations in key oncogenic drivers do not only fuel proliferation via cell intrinsic cues but also impact tumor biology via modulation of the tumor microenvironment. Along these lines, the inventors found increased pPD-L1adj. levels in patients suffering from KRAS mutated NSCLC as compared to those with KRAS wildtype status (
The inventors also explored a potential correlation of pPD-L1adj. with other clinical parameters. For example, the inventors found that patients with higher tumor stages (T, p=0.03), higher degrees of lymph node invasion (N, p=0.04) and a higher tumor grading (G, p=0.002) expressed more PD-L1 on the platelet surface (
To further elaborate on the potential of pPD-L1Adj. as a predictive biomarker in NSCLC, the inventors conducted sequential measurements of pPD-L1adj. in 12 patients undergoing conventional chemotherapy or ICI (
Subsequently, the inventors set out to probe whether the pre-therapeutically determined pPD-L1adj. level can predict the therapy response of NSCLC patients to immune checkpoint blocking antibodies. To do so the inventors analyzed the PFS of patients either treated with only conventional chemotherapy or immunocheckpoint blockade. pPD-L1 positive and negative subgroups were defined according to the median pPD-L1adj. level. In patients receiving conventional chemotherapy the inventors observed a significantly higher PFS when pPD-L1 levels were low (
Calculation and Use of pPD-L1adj. in a Clinical Setting for Predicting a Patient's Benefit from Therapy with an Immune Checkpoint Inhibitor
In
In
Human cancers are heterogeneous and biomarkers based on histopathological analyses of single tumor biopsies are often lacking robustness. Histological quantification of intratumoral PD-L1 expression is routinely performed on NSCLC biopsy material as an attempt to predict responses towards immune checkpoint inhibition, however, the correlation between expression levels and the overall response rate (ORR) is limited. In the inventors' present study they show that blood platelets are in frequent contact with lung cancer cells in vitro and in vivo and take up PD-L1 from the cancer cells in a fibronectin 1, integrin α5β1 and GPIbα dependent manner. The data provides mechanistic explanation for recent reports describing PD-L1 on platelets from patients suffering from different types of cancers. Interestingly, while it has been a paradigm that platelets degranulate and become inactive after contacting tumor cells, the inventors' data obtained in a live imaging platform suggest that platelets remain active and can re-enter circulation after interaction with tumor cells. Since pPD-L1 was not only detected on the surface of activated platelets but also in resting platelets, it is tempting to speculate on an equilibrium between intracellularly stored pPD-L1 in α-granules and cell surface pPD-L1. Indeed, a similar mechanism has been described for the uptake and redistribution of fibrinogen and immunoglobulins.
Importantly, as pPD-L1 was found to inhibit T cell function, it is likely that pPD-L1 plays a distinct role in systemic immunomodulation. Of note, pPD-L1 has recently been described in patients suffering from tumors which were classified as PD-L1 negative in biopsies. The inventors' herein presented data as well as other published studies on tumor heterogeneity suggest that immunohistochemistry-based quantification of protein expression on tissue sections from single biopsies should be interpreted with caution, as protein expression might differ spatially and temporally. Obviously, while the herein presented data suggest a highly efficient uptake of PD-L1 from lung cancer cells into platelets, it does not exclude that some pPD-L1 might be derived from other sources such as endothelial or other non-malignant cell types.
As the total blood volume is circulated up to 1000 times through the body each day, the inventors reasoned that platelets might mirror the collective PD-L1 payload of a tumor and thus might open up venues for novel biomarker strategies. In this regard it is striking that pPD-L1 not only correlated with tumor stage/grade and the occurrence of metastases but was found to be superior in predicting response towards immune checkpoint inhibition when compared to standard histological PD-L1 quantification on tumor biopsies. Since in particular lung cancer represents one of the most frequent and lethal cancers worldwide, further clinical investigation of pPD-L1 as a biomarker in NSCLC does not only hold the promise to unburden the health systems by avoiding costly and unnecessary therapies with ICI but, even more important, will avoid side effects of ICI in patients who would not benefit from this kind of therapy.
Besides the tremendous potential of pPD-L1adj. as a biomarker, the inventors believe that platelet PD-L1 might also represent a potential target for therapeutic intervention. This presumption is supported by the inventors' observation that pPD-L1 in NSCLC patients correlates with the number of T cells in TME and the number of infiltrating T cells. Similar observations in a mouse model support this finding. Along these lines it is tempting to speculate that pPD-L1 might be involved in formation of the premetastatic niche by generating an immunotolerant environment at sites distant from the primary tumor (
Last but not least, as determination of pPD-L1 is highly sensitive and pPD-L1 expression is not found in healthy individuals, pPD-L1 quantification might also be used for early detection of cancer and detection of tumor recurrence.
| Number | Date | Country | Kind |
|---|---|---|---|
| 21205516.4 | Oct 2021 | EP | regional |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/EP2022/079158 | 10/20/2022 | WO |
