Breast prostheses have been widely used in reconstructive plastic surgery and cosmetic surgery since the mid-20th century. Typically, the prostheses comprise an outer shell made of silicone elastomer filled with silicone gel or saline. The excellent features of silicone rubbers, including compatibility with organic tissues and durability, have led to the success thereof compared to other alternative materials. However, concerns have been raised about the possible biological activity thereof in case of long term implantation, in view of the different pathogenic situations reported for some specific implants.

The safety of breast prostheses, possibly related to surface structures, has been questioned since 1997 with the first case of anaplastic large cell lymphoma associated with breast prostheses (BIA-ALCL) [1 ], BIA- ALCL is a non-Hodgkin T-cell lymphoma which has been classified among hematolymphoid neoplasms by the World Health Organization (WHO) since 2017 [2, 3], A growing amount of scientific literature has attempted to assess the real epidemiology related to BIA-ALCL, although different data greatly limit the acquisition of a true estimate.

In fact, no precise data are available to date on the prevalence of BIA- ALCL in women with breast prostheses, on the type of prostheses, on the events associated with breast prostheses and on follow-up.

Up to 80% of women with BIA-ALCL have a late persistent seroma or periprosthetic effusion or seroma, which can be accompanied by breast swelling, asymmetry, capsular contracture or pain [4 - 6],

The tumor arises in the scar capsule near the macro-textured implant and in most cases remains localized, giving an excellent prognosis. The presentation of this condition appears 10.7 ± 4.6 years after the placement of the implant [7] and is diagnosed by flow cytometry performed on the fluid which accumulates between the capsule and the breast prostheses.

The first implants produced and used were characterized by a smooth surface. To prevent movements and rotations [8, 9] or capsular contractures [9-11], the use of implants with rough surfaces spread.

Evidence, albeit not definitive, connecting BIA-ALCL to macro-textured implants has raised concerns about the long-term safety thereof [12 -16], In this context, breast prostheses have been included in the list of agents with high priority for evaluation by the International Agency for Research on Cancer (IARC) for inclusion in the monographs of carcinogenic risks. A chronic inflammation present in the periprosthetic environment has also been described in the literature, mainly associated with textured prostheses, which seems to be the basis of a possible neoplastic transformation, favoring the creation of a pro-tumoral environment [17- 19], The study by Lee et al. shows a possible recurrence of breast cancer associated with textured implants based on a prospective observation of a total of 3,286 patients with breast cancer [20], Two groups were identified based on the different consistency of the positioned breast prostheses: groups of smooth and textured prostheses. The results suggested that the use of a textured implant with respect to a smooth implant can be associated with a high risk of relapse, although the exact correlation mechanism was not clearly explained in the study.

The surveys focused on cohort studies on a statistical basis and finding a possible correlation was difficult mainly due to the wide variability and uncertainties in the structure of the implant surface, producers, the implant time of residence and other factors.

Prasad et al. [21 ] describe supports in polydimethylsiloxane (PDMS) with different roughness on which fibroblasts are grown, demonstrating that the growth of fibroblasts is inversely related to roughness, thus coming to suggest that, to mitigate the appearance of fibrotic masses, it would be appropriate to increase the surface roughness of the prosthetic implants. The present invention relates to a physical model capable of reproducing the surface morphology (texture) of the external surfaces of prostheses, so as to analyze the in vitro behavior of the cells on the aforesaid surfaces.

Description

Description of the drawings

Figure 1 : (A) diagrammatic depiction of a possible embodiment of the device according to the present invention; (B, C) scanning electron microscope photographs representative of the base surface of two different subsets of wells.

Figure 2: In vitro viability test (A) Gating strategy for PBMCs (peripheral blood mononuclear cells). Physical parameters were used to exclude debris and doublets, while the vitality identification was done by staining with the fluorescent marker 7-AAD. The cells were classified as (B) viable, (C) in early apoptosis, and (D) necrotic cells. Statistical analysis **p< 0.01 (unpaired t test with Welch correction).

Figure 3: (A) Ex vivo and (B) in vitro ELISA (enzyme-linked immunosorbent assay) quantification of the indicated markers.

Figure 4: ex vivo FACS (A-Y) analysis on leukocyte samples from periprosthetic fluids. % calculated on CD4 or CD8 positive cells. Statistical analysis *p<0.05; **< 0.01 (unpaired t test with Welch correction).

Figure 5: In vitro FACS analysis (A-N) on PBMC samples grown on micro- or macro-textured surfaces. Statistical analysis *p<0.05; **< 0.01 (unpaired t test with Welch correction).

Figure 6: ELISA quantification of the markers indicated ex vivo (A-E) and in vitro (F-J). Statistical analysis *p<0.05; ** *< 0.001 (unpaired t test with Welch correction).

The present invention first relates to a multi-well plate for cell cultures, characterized in that the wells contained in said plate are grouped into at least two subsets, where the distinctive feature defining the belonging of a well to one of said subsets is the base surface roughness of said wells. The present invention further relates to an in vitro method for predicting the response of cells to surfaces with different roughness, where said method comprises:

Providing a multi-well plate comprising at least two subsets of wells characterized by a different base surface roughness;

Seeding a population of cells into said wells;

Measuring, at one or more times after seeding the cells, a biological parameter on said population of cells in said wells;

Checking any correlation between said biological parameter measured and the subset to which said well belongs, where the presence of said correlation is indicative of an effect of the surface roughness on the biological parameter under analysis.

In an embodiment, said method is used to measure the inflammatory response of mononuclear cells.

In an embodiment, said method is used to test and predict the biological compatibility and inflammatory response related to the surface of the prosthetic implant to be implanted in patients. In this embodiment, said population of cells is preferably a population of PBMC cells isolated from peripheral blood.

In an embodiment, said method is used for choosing the most suitable prosthetic implant for a given subject. In this embodiment, said population of cells is preferably a population of cells collected in an area where said prosthesis will be introduced.

In an embodiment, the surface of said wells is in PDMS (Polydimethylsiloxane, CAS number: 9016-00-6), a colorless material which allows good light transmissibility, low toxicity, low cost, biocompatible and therefore particularly suitable for in vitro cell culture. Furthermore, the prostheses are typically made of silicone materials, materials having the same nature as PDMS.

Said subset classification comprises at least two of the following types of roughness:

Smooth surface;

Micro-textured rough surface, with cavity dimensions between 10 and 80 micrometers or between 25 and 80 pm, or between 40 and 80 pm.

Macro-textured rough surface; with cavity sizes between 150 and 500 micrometers, or between 200 and 450 pm, or between 250 and 400 pm.

Said different roughnesses are obtained on the surface of said wells according to methods known to the person skilled in the art. In the situations reported, the PDMS, silicone elastomer, selected for preparing the models is Dow Corning - SYLGARD 184, a transparent, low-viscosity PDMS polymer commonly used for biological applications; other silicones with similar properties can be used.

In the event of macro-textured surfaces, 0.5 - 1.5 ml of liquid PDMS is poured into the wells of the plates. After partial cross-linking of the PDMS (30 - 50 min at T between 40 and 60 °C), appropriately sieved salt having a selected particle size and corresponding to the cavities to be obtained on the macro-textured surfaces (150-500 micrometers) is homogeneously distributed thereon. A complete covering of the PDMS surface with salt allows obtaining high-density macro-textured surfaces; a less dense covering allows obtaining low-density macro-textured surfaces. The shape and size of the cavities are a function of the shape and size of the salt crystals used. After complete cross-linking of the PDMS for 24 hours at T between 20 and 30 °C, the salt is removed and washed away so that the surfaces of the models obtained are ready for subsequent tests.

The methodology indicated above is not adapted to obtain micro-textured surfaces (10-80 micrometers); in such a case, the replication technique (coating emulation) is advantageously adopted. Starting from the surface of a pre-existing prosthesis, possibly explanted, or from a surface of adequate roughness, replicas of the micro-textured surface are made in negative using a low-viscosity thermosetting resin (e.g., epoxy resin). Resin replicas in the form of disks with a diameter equal to that of the wells are positioned covering the bottom of cylindrical containers of the same diameter. PDMS is poured into such containers and brought to cross-linking completion to obtain (positive) silicone elastomer counterreplicas in the desired number which faithfully reproduce the original surface and which are then extracted from the containers and placed on the bottom of the wells intended for subsequent tests.

The smooth surfaces are obtained by pouring the PDMS directly into the bottom of the wells and then bringing to complete cross-linking.

In an embodiment, said plate is a 24-well plate.

In an embodiment, said method is conveniently applied to the prosthesis field, where the roughness of said wells mimics the roughness of the implants to be tested.

Knowing the roughness of the prosthesis under analysis, the person skilled in the art knows how to appropriately add further subsets of wells having the desired roughness to the device.

In an embodiment, by way of non-exhaustive explanation, said prostheses are breast prostheses. In an embodiment, they are maxillary implants.

In an embodiment, said cells are cells of the immune system. By way of explanation, said cells are PBMCs (peripheral blood mononuclear cell). In an embodiment, said cells are taken from the blood of a subject; from this, the PBMCs are separated by centrifugation with Ficoll from the red blood cells and neutrophils. In this embodiment, the method is applied as a predictive tool of the inflammatory response to a prosthetic implant with said surface roughness. By obtaining immune system cells from the peripheral blood, typically responsible for the undesired reactivity following the insertion of a prosthesis, and using the same in the method according to the present invention, the data obtained is predictive of the reactivity following exposure to a prosthesis having one of the roughnesses tested, thus allowing to choose the most inert prosthesis texture.

Therefore, the present invention further relates to a method for predicting the inflammatory response related to the surface texture of a prosthesis, where said method comprises:

Providing a multi-well plate comprising at least two subsets of wells characterized by a different base surface roughness;

Seeding a population of cells in said wells, where said cells are a sample of cells isolated from a human subject;

Measuring, at one or more times after seeding the cells, a biological parameter on said population of cells in said wells;

Checking any correlation between said measured biological parameter and the subset to which said well belongs, where said correlations are predictive of the responsiveness of the human body to the implantation of a prosthesis having the roughness characterizing said subset in which the correlation was observed.

The system and method according to the present invention have been validated by comparing the expression levels of inflammatory mediators and the presence of subpopulations of immune system cells kept in culture on the plate according to the present invention with those present in the liquid present in the prosthetic space of patients. The experimental data obtained, reported in the following examples, confirm that the trend is similar; therefore, the system and the method according to the present invention are capable of reflecting what occurs in vivo, making the system and method a valid approach for checking the prosthesis before implantation, for the purposes of personalized medicine, but also to predict the inflammatory response to a prosthetic implant with said surface roughness.

The physical model provided with the present invention faithfully reproduces the surfaces, of various morphology (texture), referable to the external surfaces of breast prostheses implanted in patients, thus allowing the actuation of a method adapted to analyze the in vitro behavior of cells on the aforesaid surfaces.

The examples presented with reference to breast prostheses can be used for other types of prostheses, such as, by way of non-exhaustive explanation, pacemakers, prosthetic heart valves, vascular prostheses, orthopedic prostheses, for example hip prostheses, buttocks prostheses and dental implants.

The following examples have the sole purpose of better illustrating the invention and are not to be intended as limiting it in any manner, the scope of which is defined by the claims.

Examples

Example 1 : Preparation of a device according to the present invention The macro- and micro-textured surfaces of commercial breast prostheses were analyzed with a scanning electron microscope (SEM) to determine the main dimensional and morphological features thereof. Silicone model surfaces with selected textures were thus generated, similar to those present in the commercial implants, using methods known to the person skilled in the art. In particular, the controlled granulometry technique of covering with salt was adopted to obtain macro-textured surfaces and the replica technique to obtain microtextured surfaces.

In particular, a plate comprising two subsets of micro-wells was obtained. With reference to figure 1 , said subsets are characterized by: a microtextured base surface (FIG. 1 B), a macro-textured base surface, (Fig. 1 C).

Example 2: method validation

The base surface of said wells of desired roughness is in PDMS. Preliminarily, the effect of PDMS on cell viability was then tested. PBMC cells were seeded in cell culture wells in the absence (empty) or in the presence of PDMS, where the PDMS was of micro-texture or macrotexture roughness. By means of FACS analysis, cell viability was monitored after 2 days of culture. The data are shown in Figure 2 and clearly show how the % of living cells does not vary in the presence or absence of PDMS (Fig. 2B). Also the % of cells in early apoptosis (Fig. 2C) or necrotic apoptosis (Fig. 2D), do not undergo significant variations depending on the different surfaces.

Therefore, the experiment confirmed the absence of biological effects on the PDMS cell culture as such.

After this check, a multi-well plate was then provided, as described in example 1. It was used for the growth of PBMC cells. Two days after seeding, the following inflammation markers were measured: CCL2, IL6, IL8, CCL5. The concentrations obtained for the different markers are shown in the graphs in Fig. 3B. Note the differences observed as the base surface roughness of the well changes (black columns: macrotexture subset; white columns: micro-texture subset).

The same markers were measured in immune system cells derived from the patients' periprosthetic space. The datum shown in Fig. 3A is the average datum obtained by more than ten patients per type of prosthesis, as indicated in the figure, where some patients were exemplary of macrotexture prostheses (black columns) and others of micro-texture prostheses (white columns).

The production trend between cytokines produced in patients and that observed in the in vitro experiment is very similar.

Example 3: ex vivo study in an enlarged cohort of subjects Sample collection

Periprosthetic fluid samples were collected from a total of 37 breasts, from 29 subjects. Each sample came from intact, uninfected breast prostheses. Only samples associated with implants that had been in place for at least 6 months were used, including both prosthetic breast replacement (aesthetic or reconstructive) and second-stage breast reconstruction. The periprosthetic liquid was collected with a 10 ml sterile syringe after engraving the capsule and placed in a sterile container. FACS analysis was performed on samples obtained from 7 patients with macro-texture implants and 7 patients with micro-texture implants. An ELISA assay was performed on samples of 21 macro-textured breast prostheses and 15 micro-textured prostheses. Some of the removed prostheses were analyzed by SEM (electron scanning microscopy), to confirm the correctness of the information available regarding the texture which had been implanted. The SEM analysis always confirmed a correct match.

Flow cytometry (FACS)

Flow cytometry experiments were performed following a standard protocol and using the following monoclonal antibodies (mAbs). For the ex vivo analysis:

CD45 BV570 (Biolegend, San Diego, CA, USA; #304034); CD11 b BUV805 (BD Biosciences, Eysins, CH; #748587); CD3 BUV395 (BD Biosciences, #564001 ); CD19 APC (BD Biosciences, catalog number 555415); CD4BV421 (BD Biosciences, #562424); CD8 FITC (BD Biosciences, #555366); CCR7 PECF594 (BD Biosciences, #566768); CD45RO APC-H7 (BD Biosciences, #561137); CD95 BUV737 (BD Biosciences, #564710); CD69 PeCy7 (BD Biosciences, #57745); CD127 PeCy5 (eBioscience, San Diego, CA, USA; #15-1278-42); CD25 APCRH700 (BD Biosciences, #565106); CD56 PE (BD Biosciences, # 555516); CD16 BV786 (Biolegend, #302046); CD14 APC (BD Biosciences, #555399); CD66bPeCy7 (Biolegend; #305116); HLA-DR BV711 (BD Biosciences, #563696); CD15 BUV496 (BD Biosciences, # 741187); CD33 BUV563 (BD Biosciences, #741369); CD206 FITC (BD Biosciences, #551135); CD30 BV605 (BD Biosciences; #744408); CD163 PECF594 (BD Biosciences; #568206).

For the in vitro analysis: CD45 BV570 (Biolegend); CD11 b BLIV805 (BD Biosciences); CD3 BLIV395 (BD Biosciences); CD19 APC (BD Biosciences); CD4 BV421 (BD Biosciences); CD8 FITC (BD Biosciences); CCR7 PECF594 (BD Biosciences); CD45RO APC-H7 (BD Biosciences); CD95 BLIV737 (BD Biosciences,); CD69 PeCy7 (BD Biosciences); CD127 PeCy5 (eBioscience); CD25 APCRH700 (BD Biosciences); CD16 BV786 (Biolegend); CD14 (BD Biosciences); HLA- DR BV711 (BD Biosciences); CD206 FITC (BD Biosciences); CD30BV605 (BD Biosciences); CD163 PECF594 (BD Biosciences). The dead cells were excluded by Live and Death marking (BV510, Invitrogen #L34957). The marked cells were fixed in PBS, 1 % formalin. The acquisition was performed with FACSymphony A5 (BD Biosciences) and analyzed using FACS Diva and FlowJo software version 6.1.1 (BD Biosciences).

ELISA assay

The supernatant fluids from the cell cultures were collected after 48 hours in culture and centrifuged at 1200 rpm for 5 minutes. Fluids collected from the periprosthetic space were treated in the same manner. To quantify the production of IL-6, TNFalpha, IL-8, CCL5 and CCL2 in cellular supernatants and periprosthetic fluids, commercial ELISA kits were used following the manufacturer's instructions (R&D Systems). The data were analyzed with SoftMax Pro 5.3 software.

FACS analyses were performed on the leukocytes derived from periprosthetic fluids. The monocytes can be divided into canonical or non- canonical monocytes: the former show an inflammatory phenotype and are characterized by the surface markers CD14+/CD16+, on the contrary, the non-canonical monocytes are usually associated with an antiinflammatory environment and are characterized by a lack of CD16 on the cell surface. It has been observed that canonical monocytes decrease in the periprosthetic fluid associated with macro-textured implants as compared to the fluid associated with micro-textured implants (Figure 4a). An increase, albeit not statistically significant, of the macrophages and in particular of the more immunosuppressive subset CD163+CD206+ was observed in the macro-structured surfaces (Figure 4b-d). The other components of the immune system showed an increasing trend of CD45 positivity in eosinophils, neutrophils, natural killers (NK), CD8 cytotoxic T cells and T-regs (Figure 4e- 1).

The T cell profile revealed greater T cell maturation from the macrotexture group as compared to the micro-texture samples. In particular, a reduction in CD4+ Naive T cells was observed, associated with an increase in CM and EM type CD4+ T cells (Figure 4m- o). Similar results were reported on the CD8+ T cells as evidenced by the shrinkage of the naive subgroup (Figure 4q-s).

In patients implanted with the macro-texture prosthesis, an increase in CD69-positive cells was observed (Figure 4u-v and x-y), despite not achieving statistical significance, while HLA-DR and CD30- positive cells were found to be at a similar level (Figure 4 p, t and w).

Taken together, these data indicate that the microenvironment of the macro-texture prosthesis is more immunosuppressive than that of the micro-texture prosthesis.

Human peripheral blood mononuclear cells (PBMCs) cultured on PDMS replicates of micro-texture and macro-texture surfaces for 48 hours without vital stimuli were analyzed in parallel with FACS to characterize the immune activation. A decrease in monocytes (Figure 5A) and an increase in T cells (Figure 5B) were observed on macro-structured surfaces as compared to what observed on micro-structured surfaces, No variation was observed in the CD4 and CD8 subsets (Figure 5C-D). As observed in the ex vivo assay, an increase in CD4 in CM and EM cells was observed in the in vitro model (Figure 5 j-k) and two CD8 donors (Figure 5H-I). Conversely, naive CD4 cells and terminal CD4 effectors, and CD8 cells showed no variation (Figure 5E) while naive CD8 is reduced on macro-structured surfaces (Figure 5I). Activated CD4 (Figure 5F) and CD8 (Figure 5N) cells, positive for CD69+ were increased in the derived PBMCs from all donors analyzed cultured on macrotextured surfaces, similarly to what was observed ex-vivo (Figure 4u-v and x-y).

In order to check whether the periprosthetic microenvironment of the patients was related to texture, several cytokines were studied. In particular, IL6 and IL8, two pro-inflammatory cytokines, typically increased in the tumor microenvironment, TNF-alpha, a cytokine associated with T-cell activation, and two chemokines, CCL2 and CCL5, which can recall monocytes or lymphocytes, respectively. The analysis of the soluble mediators in the periprosthetic fluid revealed that levels of IL6, TNF alpha and IL8 were significantly higher in the macro-textured implants than in the micro-structured ones (Figure 6 a-c), while CCL5 and CCL2 are similarly reduced (Figure 6 d-e). Consistent with what was found in the periprosthetic space, soluble mediators released by PBMCs culturedon macro-texture surfaces were higher than cells cultured on micro-texture surfaces (Figure 6 f-h); furthermore, in this model CCL5 is released at the same level, while CCL2 appears to be less expressed in the PBMCs cultured on macro-texture (Figure 6 i and j).

The data collected indicate that macro-texture surfaces activate an immune cell response and can cause an inflammatory microenvironment.This datum obtained from patients' periprosthetic fluid is replicated following the method of this patent and can thus be used to predict the inflammatory response to a prosthetic implant with said surface roughness.

References

1. Keech et al Anaplastic T-cell lymphoma in proximity to a saline- filled breast implant. PRS doi:10.1097/00006534-199708000- 0006.

2. Srinivasa et al. Global Adverse Event Reports of Breast Implant- Associated ALCL: An International Review of 40 Government Authority Databases. PRS doi:10.1097/PRS.000000000000323

3. Arber et al. The 2016 revision to the World Health Organization classification of myeloid neoplasms and acute leukemia. Blood doi: 10.1182/blood-2016-06-721662

4. Marra et al. Breast implant-associated anaplastic large cell lymphoma: A comprehensive review. Cancer treatment reviews doi:10.1016/j.ctrv.2020.101963

5. Clemens et al. Complete Surgical Excision Is Essential for the Management of Patients With Breast Implant-Associated Anaplastic Large-Cell Lymphoma. J Clin Oncol, doi: 10.1200/JC0.2015.63.3412

6. Mehta-Shah et al. How I treat breast implant-associated anaplastic large cell lymphoma. Blood doi: 10.1182/blood-2018- 03-785972

7. Doren et al. U.S. Epidemiology of Breast Implant-Associated Anaplastic Large Cell Lymphoma. PRS doi: 10.1097/PRS.0000000000003282

8. Maxwell et al. Natrelle style 410 form-stable silicone breast implants: core study results at 6 years. Aesthetic surgery journal doi: 10.1177/1090820X12452423

9. Maxwell et al. Benefits and Limitations of Macrotextured Breast Implants and Consensus Recommendations for Optimizing Their Effectiveness. Aesthetic surgery journal doi: 10.1177/1090820X14538635

10. Barnsley et al. Textured surface breast implants in the prevention of capsular contracture among breast augmentation patients: a meta-analysis of randomized controlled trials. PRS doi: 10.1097/01 .prs.0000218184.47372.d5

11. Wong et al. Capsular contracture in subglandular breast augmentation with textured versus smooth breast implants: a systematic review. PRS doi: 10.1097/01 .prs.0000237013.50283.d2

12. Marra et al. Breast implant-associated anaplastic large cell lymphoma: A comprehensive review. Cancer treatment reviews doi:10.1016/j.ctrv.2020.101963

13. Clemens et al. Complete Surgical Excision Is Essential for the Management of Patients With Breast Implant-Associated Anaplastic Large-Cell Lymphoma. J Clin Oncol, doi: 10.1200/JC0.2015.63.3412

14. Mehta-Shah et al. How I treat breast implant-associated anaplastic large cell lymphoma. Blood doi: 10.1182/blood-2018- 03-785972

15. Doren et al. U.S. Epidemiology of Breast Implant-Associated Anaplastic Large Cell Lymphoma. PRS doi: 10.1097/PRS.0000000000003282

16. IARC Monographs Priorities Group. Advisory Group recommendations on priorities for the IARC Monographs. The Lancet. Oncology doi: 10.1016/S1470-2045(19)30246-3

17. Salvans et al. Postoperative peritoneal infection enhances migration and invasion capacities of tumor cells in vitro: an insight into the association between anastomotic leak and recurrence after surgery for colorectal cancer. Annals of surgery doi: 10.1097/S LA.0000000000000958

18. Dorcaratto et al. Impact of Postoperative Complications on Survival and Recurrence After Resection of Colorectal Liver Metastases: Systematic Review and Meta-analysis. Annals of surgery doi: 10.1097/SLA.0000000000003254

19.Aoyama et al. Impact of postoperative complications on survival and recurrence pancreatic cancer. Anticancer research vol. 35,4: 2401-9. 20. Lee et al. Association of the Implant Surface Texture Used in

Reconstruction With Breast Cancer Recurrence. JAMA surgery doi:10.1001/jamasurg.2020.4124