METHOD FOR PREDICTING CANCER PROGRESSION BY NANOMECHANICAL PROFILING

Abstract

The invention relates to a method for staging metastatic potential of a primary tumour sample or related lymph node by nanomechanical measurement and/or for determing the reoccurrence or incidence potential.

Claims

1. A method for classifying a tissue sample obtained from a patient, wherein said tissue sample is suspected to comprise secondary tumour tissue, said method comprising determining a stiffness value for each of a first plurality of points on a primary tumour sample, resulting in a first stiffness distribution, determining a stiffness value for each of a second plurality of points on said tissue sample, resulting in a second stiffness distribution, wherein said tissue sample is classified as a metastasis if said first stiffness distribution and said second stiffness distribution both show a heterogeneous stiffness distribution having a frequency maximum at substantially the same stiffness value below 1 kPa.

2. The method of claim 1, wherein both said first and said second pluralities of points are determined with a spatial resolution of at least 100 μm.

3. The method of any one of claim 1 or 2, wherein said tissue sample was taken from a lymph node, particularly adjacent to the sampling site of said tumour biopsy sample or an axillary lymph node.

4. A method for classifying a tissue sample obtained from a tumour, comprising determining a stiffness value for each of a plurality of points on said sample with a spatial resolution of at least 100 μm, resulting in a stiffness distribution, assigning a probability of malignancy to said sample, wherein said method is applied to a first sample and a second sample, said first sample is a primary tumour sample, and said second sample is a sample taken from a lymph node, particularly a lymph node adjacent to said sampling site of said first sample or an axillary lymph node, and said second sample is classified as a lymph node metastasis if said first sample and said second sample both show a heterogeneous stiffness distribution having a frequency maximum below 1 kPa, and the frequency maximum of the second sample is the same as the frequency maximum of the first sample.

5. A method for classifying a tissue sample obtained from a tumour, comprising determining the stiffness values for a plurality of points on said sample with a spatial resolution of at least 100 μm, resulting in a stiffness distribution, assigning to said sample to a probability of malignancy, wherein a first stiffness distribution is obtained from a first site of said sample and a second stiffness distribution is obtained from a second site of said sample, and said first site corresponds to a part of said tumour histologically classified as tumour tissue and said second site corresponds to adjacent tissue, particularly tissue histologically classified as beyond the border of the tumour, and said tumour sample is classified as having a low probability of having spread to said adjacent tissue, particularly to adjacent lymph nodes or an axillary lymph node, if said first stiffness distribution is a heterogeneous stiffness distribution having a frequency maximum below 1 kPa, and said second frequency distribution is characterized by absence of a stiffness distribution frequency maximum below 1 kPa, and/or said tumour sample is classified as having a high probability of having spread to said adjacent tissue, particularly to adjacent lymph nodes or an axillary lymph node, if said second frequency distribution is characterized by presence of a stiffness distribution frequency maximum below 1 kPa.

6. The method according to any one of the preceding claims, wherein said tissue sample is a tissue biopsy sample or a resection specimen.

7. The method according to any one of the previous claims, wherein said plurality of points is arranged as a grid of n.sub.1 by n.sub.2 points, said grid defining an area.

8. The method according to any one of the previous claims, whereby said stiffness values of at least two different areas of said same sample are determined, and the distance between the geometrical centres of said areas is a multiple of said spatial resolution of at least 10.

9. The method according to any one of the preceding claims, wherein said plurality of points comprises 100, 400, 900, 1000, 1600, 2500, 3600, 4900, 6400, 8100 or 10000 stiffness values.

10. The method according to any one of the previous claims, characterized by that said tissue sample is a cylindrical or prismatic biopsy with a diameter of at least 7 μm.

11. The method according to any one of the previous claims, wherein said tumour is a human mammary carcinoma or a lymph node, lung, bone, liver or brain metastasis.

12. The method according to any one of the previous claims, characterized by that said stiffness values are determined under physiological conditions.

13. The method according to any one of the preceding claims, wherein a primary tumour sample exhibiting a frequency maximum below 0.5 kPa is classified as metastasized tumour or as having a high probability of having spread to adjacent tissue, particularly to adjacent lymph nodes or to axillary lymph nodes.

14. A method for staging cancer, comprising obtaining a first tissue sample from a primary tumour and a second tissue sample, determining the stiffness values for a plurality of points on said first tissue sample and for a plurality of points on said second biopsy sample, with each of said plurality of points being characterized by a spatial resolution of at least 100 μm, resulting in a stiffness distribution for each of said first tissue sample and said second tissue sample, assigning to said first sample a probability of malignancy, and assigning to said second sample a probability of being invaded by said primary tumour.

15. The method according to claim 14, wherein said second tissue sample is obtained from a tissue adjacent to said primary tumour or a lymph node, particularly an adjacent lymph node or an axillary lymph node.

16. The method according to claim 14 or 15, wherein said first tissue sample and/or said second tissue sample is a tissue biopsy sample or a resection specimen.

17. The method according to any one of claims 14 to 16, wherein said first sample is assigned a high probability of being malignant if said first sample is characterized by an at least bimodal stiffness distribution having a first peak exhibiting an at least two-fold higher stiffness value than a second peak, and/or said second sample is assigned a high probability of being invaded by said primary tumour if said second sample is characterized by an at least bimodal stiffness distribution characterized by a first peak exhibiting an at least two-fold higher stiffness value than a second peak.

18. The method according to any one of claims 14 to 17, wherein said tissue adjacent to said primary tumour is comprised within a lymph node.

19. The method according to any one of claims 14 to 18, wherein a first sample exhibiting a second peak below 0.5 kPa is classified as metastasized tumour.

20. A method for staging cancer, comprising obtaining a tissue sample form a tissue adjacent to a primary tumour, determining the stiffness values for a plurality of points on said sample with a spatial resolution of at least 100 μm, resulting in a stiffness distribution, assigning to said sample a probability of being invaded by said primary tumour.

21. The method according to claim 20, wherein said tissue sample is a tissue biopsy sample or a resection specimen.

22. The method according to claim 20 or 21, wherein said sample is assigned to a high probability of being invaded by said primary tumour if said stiffness distribution is characterized by a maximum between 0.2 kPa and 1 kPa, particularly between 0.3 kPa and 0.8 kPa.

23. The method according to any one of claims 20 to 22, wherein a sample showing an at least bimodal stiffness distribution is assigned a high probability of being invaded by said primary tumour, wherein said at least bimodal stiffness distribution is characterized by a first peak exhibiting an at least two-fold higher stiffness value than a second peak.

24. A method for classifying a tissue sample obtained from a tumour comprising determining the stiffness values for a plurality of points on said sample with a spatial resolution of at least 100 μm, resulting in a stiffness distribution, assigning to said sample to a probability of malignancy, wherein a sample exhibiting a peak in said stiffness distribution below 0.5 kPa is classified as metastasized tumour.

25. The method according to claim 24, wherein said tissue sample is obtained from a mammary carcinoma.

26. The method according to claim 24 or 25, wherein said tissue sample is a tissue biopsy sample or a resection specimen.

27. The method according to any one of the preceding claims, wherein said primary tumour or said tumour is a mammary carcinoma, kidney tumour, prostate tumour, brain tumour, lung tumour, ovarian tumour, pancreas tumour, or stomach tumour, liver tumour, skin tumour or gastric tumour.

28. A device for tumour sample diagnosis, comprising an atomic force microscope and a computer connected thereto, the computer being configured to run a programme conducting the method of any of the previous claims.

29. The method according to any one of the preceding claims, wherein said spatial resolution is 20 μm, 10 μm, 5 μm or 1 μm.

Description

[0123] The invention is further illustrated by the following examples and figures, from which further embodiments and advantages can be drawn. These examples are meant to illustrate the invention but not to limit its scope.

[0124] Short description of the figures

[0125] FIG. 1 shows A—tumour B—border of the tumour and adjacent tissue C—adjacent tissue.

[0126] FIG. 2 shows the soft nanomechanical profile of the primary tumour and the adjacent tissue reveal a common phenotype detected in the lymph node metastases. (A) Nanomechanical profiles of two cancer patients reveal bimodal distributions with an exponential decay in the case of patient 1 with peak values at 0.39+/−0.21 kPa for specific cancer cell population and 3.66+/−2.14 kPa decaying up to 10 kPa for the surrounding cellular and stromal components of the tissue. Adjacent tissue with pathohistological characteristics of normal breast tissues exhibits a uniform stiffness peak of 1.9+/−0.7 kPa (top). Sharp distribution and overall softer phenotype is measured for patient 2 with peak values for cells at 0.45+/−0.18 kPa and 1.0 +/−0.55 kPa respectively (bottom). Adjacent tissue of this patient shows a bimodal distribution with a very prominent soft peak of 0.4+/−0.2 kPa as well a stiffer peak of 1.05+/−0.5 kPa. In the pathohistological analysis this tissue is marked as non-malignant. Therefore here we use the term “Corrupted” healthy since this tissue stiffness values and bimodal distribution do not correspond to the values measured for healthy breast samples from control patients. (B) Patient 1 has cancer negative lymph node exhibiting stiffness values of 3.02+/−2.10 kPa (top). Lymph node containing breast cancer metastases from patient 2 exhibits bimodal stiffness distribution with peak values for cancerous region of 0.49+/−0.23 kPa that corresponds directly to “soft” cancer cell phenotype found in the primary tumour and the adjacent (“Corrupted” Healthy tissue) while the value of 2.11+/−0.78 kPa is measured for the rest of the lymph node (bottom).

[0127] FIG. 3 shows a post-AFM histological overview of the resection specimens from two breast cancer patients. H&E staining of both samples exhibit similar features; i.e. connective tissue and normal epithelium not significantly different to healthy breast (top). Primary tumours in both cases reveal an invasive breast carcinoma cells with infiltrating nests of cells that have evoked a dense fibrous tissue response (middle). H &E staining of the corresponding lymph node from the patient 2 reveals nests of breast cancer cells similarly to the primary tumour (bottom right), while the lymph node from patient 1 is clear of cancer cells (bottom left).

[0128] FIG. 4 shows normalized histograms representing the nanomechanical profile of the primary tumours (biopsy samples and resection specimens) of patient without invaded lymph node N0 (n=10) and with invaded lymph nodes N+(n=11). N+ patients clearly show a shift to softer values in the cancer area (below 1 kPa). They peak at 0.325 kPa, where the NO patients peak at 0.625 kPa. Both distributions remain quite broad and the part of the histogram above 2 kPa looks very similar in both cases.

[0129] FIG. 5 shows normalized histograms representing the nanomechnical profile non-invaded and invaded lymph nodes corresponding to the primary tumours illustrated in FIG. 4. Cancer cell positive (N+) lymph nodes have a strong, soft peak around 0.4 kPa, whereas cancer negative (N−) lymph nodes lack a soft peak.

EXAMPLES

[0130] In order to transfer preliminary results into a clinical setting, the ARTIDIS technology was optimized for analysis of unfixed (measured in physiological aqueous environment or frozen tissue) human breast cancer samples obtained by tumour resections. For this purpose, 152 tissue samples, including primary breast cancers of various stage and grade, lymph node metastases, and non-neoplastic human breast tissues were collected from resection specimens of 56 patients undergoing either lumpectomy or mastectomy procedures.

[0131] Nanomechanical measurements of the samples were performed as disclosed in U.S. Pat. No. 8,756,711 B2. Briefly, each sample was examined in a systematic manner by homogeneously distributing FV maps over the whole sample surface to account for possible heterogeneities. A regular distance of approximately 500 μm was kept between the scan using either micrometer screws or automated positioning systems. This resulted in roughly 10 to 15 FV maps per specimen depending on the total biopsy size.

[0132] For the analysis of the samples by AFM, biopsies were glued onto a culture dish using 2-component 5-minute fast drying epoxy glue. After a pre-drying step of 2 minutes (to avoid mixing of the epoxy and the specimen buffer), the specimen was laid flat onto the glue in order to optimize the indentation angle and to avoid influence from external components (e.g. the cantilever holder). Pipette tips acting as “ramps” were placed directly under uneven segments of each specimen to maintain height consistency. The use of excessive force (e.g. tearing or stretching) was minimized at all times during specimen handling. All preparative steps were performed in either a sterile buffer environment supplemented with protease inhibitors or transplantation buffer to prevent contamination and to ensuring that the specimen remained in a close-to-in-vivo state. The mounted specimens were kept in ice-cold Ringer's solution or Custodiol until nanomechanical testing, which was performed at room temperature or at 37° C.

[0133] For sharp pyramidal tips (205-μm-long silicon nitride cantilevers, nominal cantilever spring constant k=0.06 N m.sup.−1, resonance frequency [air]=18 kHz), the exact spring constant k of the cantilever was determined prior to every experiment with the thermal tune method while the deflection sensitivity was determined in fluid using solid glass substrates as an infinitely stiff reference material.

[0134] Contact stiffness (elastic modulus, E) measurements of biopsies were derived as follows; load-displacement curves, also designated as force indentation curves, were recorded at a given site in an oriented manner during both loading and unloading. A regular distance of approximately 500 μm was kept between the scan regions using either micrometer screws or automated positioning systems. An individual set of data consisted of 1,024 load-displacement curves, at an indentation speed of 16 μm/s. This resulted in roughly 15 to 20 force volume maps per sample. When possible, force- volume maps (FV) were made over a 32×32 point grid with a scan size of 20×20 μm at a rate of approx. 1 load and unload cycles per second. Each load-displacement curve consisted of at least 512 data points whereas the Z length was set to 5 μm to 8 μm depending on the properties of the analyzed region. Each FV map was set to 20×20 μm.sup.2 in order to (i) optimize experimental time as well as (ii) to provide a sufficiently large area incorporating all components within the tissue (e.g., cells and extracellular matrix). The maximum applied loading force was set to 1.8 nN and an indentation depth of approximately 150 to 3000 nm. Additional 72×72 FV maps (5184 force-displacement curves per map and a pixel size of 277 nm) were obtained to increase the spatial resolution over key areas of interest.

[0135] Force indentation curves were analyzed using a method described previously (Loparic, et al., Biophysical Journal, 98(11): p. 2731-40, 2010, Plodinec, et al., Journal of Structural Biology, 174(3): p. 476-484, 2011). Briefly, software was developed in LabVIEW (National Instrument, US) for the automated analysis of the FV data. The contact point was determined. Force curves were obtained transforming from piezo displacement to tip-sample distance, which accounts for the bending of the cantilever and by multiplying cantilever deflection d with the spring constant k to obtain the load F. Unloading force curves were analyzed by performing a linear fit to the upper 50% of the force curve, which defines the stiffness between the maximum load F=1.8 nN and a load of 0.9 nN. Extraneous effects on the force curve such as adhesion could be avoided by this procedure. The Poisson ratio was set to 0.5. The Young's modulus was determined according to the Oliver and Pharr method (Oliver et al., Journal of Materials Research, 7(6), 1564-1583, 1992). The slope values were spatially plotted, analyzed and displayed in ARTIDIS OFFLINE SOFTWARE.

[0136] Post-AFM, tissue samples were fixed and paraffin embedded in an oriented manner. ARTIDIS data have confirmed the initial findings that all carcinoma samples display heterogeneous stiffness phenotypes with a characteristic 2-fold softer phenotype in comparison to the surrounding non-neoplastic and morphologically normal breast tissue. Healthy mammary tissue of patients without breast cancer exhibits on average stiffness values ˜1.6 kPa.

[0137] Most importantly, the data illustrated in FIG. 2 and FIG. 3 show:

[0138] 1) Tumour tissues from breast cancer patients exhibit a heterogeneous distribution from 0.4 kPa and an exponential decay that can range up to 20 kPa. We identified invasive breast cancer specimens by a characteristic soft peak of 0.4 to 0.8 kPa. Stiffness distribution of corresponding lymph node metastases from a same patient was characterized by a heterogeneous stiffness profile with a characteristic soft peak of 0.4 to 0.8 kPa similarly to the primary breast cancer tissue.

[0139] 2) Adjacent tissue of these patients that was histologically rated as “non-malignant” presented a bimodal distribution with prominent soft stiffness peak ranging from 0.4 to 0.8 kPa. The presence of such tissue that according to nanomechanical analysis is cancerous, but histologically is non-malignant, is an indicator of poor prognosis. In addition stiffness values from 1.2 to 1.9 kPa that correspond to the healthy breast tissue were present as well.

[0140] 3) In cases where patients had soft stiffness peak detected only in the primary tumour, but no soft peaks in the adjacent tissue, no lymph node metastases were present.

[0141] 4) In case when fat tissue is measured, specific stiffness peak of 0.2 kPa is present.

[0142] 5) Because of usual increase of fat component within the breast tissue during aging, very often a fat specific stiffness peak of 0.2 kPa is measured. Additionally, the data illustrate in FIGS. 4 and 5 that tumours that have already spread to adjacent lymph nodes show a shift to softer values in the cancer area.

[0143] Accordingly, for assessment of cancer aggressiveness and prognosis in breast cancer patients it is important to take into account not just the primary tumour but also the nanomechanical response of the adjacent tissue.

[0144] The data presented herein demonstrate applicability of nanomechanical profiling using ARTIDIS in clinics for:

[0145] 1) Prognosis of cancer incidence, progression and recurrence

[0146] 2) Prediction of the treatment response

[0147] 3) Deciding on the appropriate treatment and follow up regimen based on the combined nanomechanical profile of primary tumour and the adjacent tissue

[0148] 4) Screening of the histologically “non-malignant” tissue (i.e. non-malignant breast tissue with no obvious pathological changes observed with standard screening methods such as ultrasound, mammography or H&E staining after a local biopsy is performed) for the soft stiffness peak with purpose of identifying presence of locally invasive cancer cells and aggressiveness degree of the tumour long before solid tumour growth and presence of symptoms

[0149] 5) Screening using ARTIDIS nanomechanical profiling is particularly suitable for patients carrying genetic mutations such as BRCA1 and BRCA2 or any other, who are at high risk for cancer development

[0150] The nanomechanical profiling method of the invention is ideally suited for use in daily practice as it allows fast, on-site assessment of specimen and does not suffer from inter-observer variability as for example other markers, such as Ki-67.

[0151] The method of the invention is based on the following principles:

[0152] 1. Tissues (cells and extracellular matrix) undergo mechanical/structural alterations at the nanometer scale before tumour (cancer) starts to develop.

[0153] 2. Alterations from Point 1 are present across the organ or adjacent tissue (e.g. diffuse/multifocal appearance).

[0154] 3. The exact location where cancer will first occur depends on the specific local micro-environmental conditions (e.g. alteration and interactions within cells and surrounding extracellular matrix and in-between them).

[0155] 4. Measurement of nanomechanical profile of the organ/adjacent tissue can detect specific alterations from Point 1.

[0156] 5. Measurement of nanomechanical profile of the organ/adjacent tissue can distinguish between age related mechanical alteration and pre-tumour, tumour and inflammation related alterations.

[0157] 6. Measurement of nanomechanical profile of the organ/adjacent tissue can correlate different types and/or degrees of mechanical alterations with tumour ability to progress and/or seed metastases.

[0158] 7. Results of Point 6 can be used as prognostic and/or predictive marker of tumour development and progression. This can be practically used for better treatment of patients for specific disease.

[0159] 8. The methodology of the invention is typically applied for detection of breast tumours (e.g. screening) and as prognostic and marker of already existing tumour or tumour associated metastases but is not limited to breast tissue. It can be also used for other organ specific tumours, inflammatory or medically relevant disease like prostate cancer, colorectal cancer, lung cancer, malignant melanoma, liver cancer, dermatitis, gastritis etc.

[0160] 9. By using the proposed methodology specific conditions: dysplasia, metaplasia, hyperplasia (pre-tumor changes) and age related changes can be specifically detected and distinguished from tumour related changes.