BCR-ABL inhibitors for use in treating keloid disease

Abstract

A breakpoint cluster region-Abelson (Bcr-Abl) inhibitor for use in a method of treating a skin condition or disease, such as keloid disease (KD) or folliculitis keloidalis nuchae (FKN), is provided The Bcr-Abl inhibitor can be GNF-2, GNF-5 or ON012380. Optionally, the Bcr-Abl inhibitor can be administered in combination with one or more 4-anilinoquinazolines, such as dasatinib, nilotinib, bosutinib, axitinib, ponatinib, rociletinib, vemurafenib, dabrafenib, trametinib, cobimetinib, binimetinib, lapatinib, sunitinib, sorafenib or pazopanib. A UDP- glucuronosyltransferase (UGT) inhibitor, such as magnolol, can also be administered in an optional second treatment step, typically about 2 to 4 weeks after the first treatment. One or more 4-anilinoquinazolines can also be administered in the second treatment step. A Bcr-Abl inhibitor and / or one or more 4-anilinoquinazolines can further be administered in an optional third treatment step, typically about 6 weeks after the first treatment. A method of treatment and a kit are also provided.

Description

[0001] BCR-ABL INHIBITORS FOR USE IN TREATING KELOID DISEASE

[0002] CROSS-REFERENCE TO RELATED APPLICATIONS

[0003] This application claims priority to Australian provisional patent application no. 2023902059 filed on 29 June 2023, which is incorporated by reference herein.

[0004] FIELD

[0005] A method and compounds for treating keloid disease are disclosed.

[0006] BACKGROUND

[0007] Keloid disease (KD) comprises benign fibroproliferative dermal tumours of unknown aetiology that can occur as a result of trauma to the human skin. Unlike other forms of scarring, KD does not regress spontaneously and the keloids (also referred to as keloid scars) grow beyond the margins of the original wound, “invading” adjacent healthy normal skin.

[0008] Clinically, KD scars are often characterized by severe itching and pain, particularly when inflamed. Anatomically, KD scars usually occur on the sternum, earlobes, shoulders, upper arm, and upper back. They can cause physical impairment and be cosmetically unsightly which may present with psychosocial stress for KD patients.

[0009] KD therapeutic management and recommended treatment modalities are ill-defined. Treatment typically includes one or more of cortisone or steroid injections, freezing with liquid nitrogen, and surgical excision. However, the outcomes of these treatments are often unsatisfactory and keloids frequently recur following surgery, sometimes growing even larger than before. Lack of relevant study models (including a lack of an animal model as keloids only occur in human skin) is one of the challenges in the development of effective therapies.

[0010] There is therefore a need for a more effective treatment for KD. SUMMARY

[0011] According to a first aspect of the disclosure, there is provided a breakpoint cluster region-Abelson (Bcr-Abl) inhibitor for use in a method of treating a skin condition or disease in a subject.

[0012] The skin condition or disease may be keloid disease (KD). Additionally or alternatively, the skin condition or disease may be folliculitis keloidalis nuchae (FKN).

[0013] The Bcr-Abl inhibitor may be GNF-2 or an analog or derivative or pharmaceutically acceptable salt thereof. The Bcr-Abl inhibitor may be ON012380. The analog of GNF-2 analog may be GNF-5.

[0014] The Bcr-Abl inhibitor may be administered in combination with a 4-anilinoquinazoline.

[0015] The 4-anilinoquinazoline may be selected from the group consisting of dasatinib, nilotinib, bosutinib, axitinib, ponatinib, rociletinib, vemurafenib, dabrafenib, trametinib, cobimetinib, binimetinib, lapatinib, sunitinib, sorafenib and pazopanib, or a derivative or pharmaceutically acceptable salt thereof.

[0016] The 4-anilinoquinazoline may be selected from the group consisting of dasatinib, nilotinib, bosutinib and axitinib, or a derivative thereof. Alternatively, the 4-anilinoquinazoline may be a pan-Bcr-Abl inhibitor selected from ponatinib and rociletinib, or a derivative thereof. Alternatively, the 4- anilinoquinazoline may be a MARK pathway inhibitor selected from the group consisting of vemurafenib, dabrafenib, trametinib, cobimetinib and binimetinib or a derivative thereof. Alternatively, the 4-anilinoquinazoline may be a P13K-Akt inhibitor selected from idelalisib and ibrutinib. Alternatively, the 4-anilinoquinazoline may be a multi receptor tyrosine kinase inhibitor (mRTKI) selected from the group consisting of lapatinib, sunitinib, sorafenib and pazopanib or a derivative thereof.

[0017] The 4-anilinoquinazoline may be nilotinib.

[0018] The Bcr-Abl inhibitor and 4-anilinoquinazoline may be administered at the same time or sequentially.

[0019] The Bcr-Abl inhibitor and 4-anilinoquinazoline may be administered to the subject shortly before, during or shortly after keloid surgery.

[0020] The method of treatment may further comprise administering a UDP-glucuronosyltransferase (UGT) inhibitor to the subject. The UGT inhibitor may be magnolol or a derivative or salt thereof. The UGT inhibitor may be administered about one to about four weeks after administration of the Bcr-Abl inhibitor.

[0021] The method of treatment may further comprise administering a 4-anilinoquinazoline to the subject about one to four weeks after the Bcr-Abl inhibitor was first administered.

[0022] The UGT and / or 4-anilinoquinazoline may be administered about 2 weeks after administration of the initial Bcr-Abl inhibitor.

[0023] The UGT and / or 4-anilinoquinazoline may be administered about 4 weeks after administration of the initial Bcr-Abl inhibitor.

[0024] The method of treatment may further comprise administering one or more 4-anilinoquinazolines to the subject about six weeks after administration of the initial Bcr-Abl inhibitor.

[0025] A 4-anilinoquinazoline that is administered about six weeks after administration of the initial Bcr-Abl inhibitor may be selected from the group consisting of vemurafenib, dabrafenib, trametinib, cobimetinib and binimetinib, or a derivative thereof.

[0026] A 4-anilinoquinazoline that is administered about six weeks after administration of the initial Bcr-Abl inhibitor may be selected from the group consisting of idelalisib or ibrutinib, or a derivative thereof.

[0027] A 4-anilinoquinazoline that is administered about six weeks after administration of the initial Bcr-Abl inhibitor may be selected from the group consisting of lapatinib, sunitinib, sorafenib and pazopanib, or a derivative thereof.

[0028] A Bcr-Abl inhibitor selected from the group consisting of GNF-2, GNF-5 and ON012380, or an analog or derivative thereof, may be administered with the 4-anilinoquinazoline that is administered about six weeks after administration of the initial Bcr-Abl inhibitor.

[0029] According to a second aspect of the disclosure, there is provided a combination of a Bcr-Abl inhibitor and a 4-anilinoquinazoline for use in a method of treating a skin condition or disease. The Bcr-Abl inhibitor may be selected from the group consisting of GNF-2, GNF-5 and ON012380. The 4- anilinoquinazoline may be selected from the group consisting of dasatinib, nilotinib, bosutinib, axitinib, ponatinib, rociletinib, vemurafenib, dabrafenib, trametinib, cobimetinib, binimetinib, lapatinib, sunitinib, sorafenib and pazopanib, or a derivative thereof. More particularly, the Bcr-Abl inhibitor may be GNF-2 and the 4-anilinoquinazoline may be nilotinib. According to a third aspect of the disclosure, there is provided a composition including:

[0030] (i) a Bcr-Abl inhibitor which is not a 4-anilinoquinazoline; and

[0031] (ii) a 4-anilinoquinazoline; for use in a method of treating a skin condition or disease. The Bcr-Abl inhibitor which is not a 4- anilinoquinazoline may be selected from the group consisting of GNF-2, GNF-5 and ON012380. The 4-anilinoquinazoline may be selected from the group consisting of dasatinib, nilotinib, bosutinib, axitinib, ponatinib, rociletinib, vemurafenib, dabrafenib, trametinib, cobimetinib, binimetinib, lapatinib, sunitinib, sorafenib and pazopanib, or a derivative thereof. More particularly, the Bcr-Abl inhibitor may be GNF-2 and the 4-anilinoquinazoline may be nilotinib. The composition may be a pharmaceutical composition.

[0032] According to a further aspect of the disclosure, there is provided a use of a Bcr-Abl inhibitor in the manufacture of a medicament for treating a skin condition or disease. The Bcr-Abl inhibitor may be a compound as previously described and may be used as described above.

[0033] According to a further aspect of the disclosure, there is provided a method of treating a skin condition or disease in a subject, the method comprising administering an effective amount of a breakpoint cluster region-Abelson (Bcr-Abl) to the subject. The method of treatment may be substantially as described above.

[0034] The method of treating keloid disease may further include a step of surgically removing a KD scar from the subject prior to or at the time of administering the first treatment.

[0035] The method of treatment may further include the steps of obtaining a sample of a keloid scar from the subject and detecting EPHA8, MAPK12, DYRK2, MAPK11 , PRKD1 , PRKCQ, PAK1 , PRKG1 , MAP4K4, LYN, AKT3, IRAKI , JAK3, PRKACB, NEK6, BMPR1 B, FGFR1 , ABL2, IGF1 R, CAMK2A, UGT1A6, UGT1A7, SERPINH1 , ANGPTL2, UGT1A1 , UGT2A1 , UGT1 A10, UGT1 A3, UGT1 A9, and UGT1A8, or any combination thereof, in the sample, prior to administering the first treatment. More particularly, the biomarkers are EPHA8 and / or MAPK12.

[0036] The method of treatment may further include the steps of obtaining a sample of a keloid scar from the subject and detecting that OMD, FAP, RCN1 , TNC, COL12A1 , CALD1 , THBS2 / TSP2, COL6A2 and PDIA5, or any combination thereof, are overexpressed relative to a sample from an adjacent normal skin region. According to a further aspect of the disclosure, there is provided a kit comprising:

[0037] (i) a Bcr-Abl inhibitor which is an allosteric inhibitor, e.g. GNF-2, GNF-5 or ON012380;

[0038] (ii) at least one 4-anilinoquinazoline selected from dasatinib, nilotinib, bosutinib, axitinib, ponatinib, rociletinib, vemurafenib, dabrafenib, trametinib, cobimetinib, binimetinib, lapatinib, sunitinib, sorafenib and pazopanib, or a derivative thereof;

[0039] (iii) a UGT inhibitor, e.g. magnolol;

[0040] (iv) instructions for treating a patient with a skin condition or disease according to the method described above.

[0041] More particularly, the Bcr-Abl inhibitor may be GNF-2 and the 4-anilinoquinazoline may be nilotinib.

[0042] According to a further aspect of the disclosure, there is provided a method of treating keloid disease in a human subject, the method comprising the steps of: a) administering a first treatment to the subject, the first treatment comprising a 4- anilinoquinazoline, optionally in combination with an allosteric inhibitor, wherein the 4- anilinoquinazoline is selected from: i) a Bcr-Abl inhibitor; ii) a MARK pathway inhibitor; iii) a PI3K-Akt inhibitor; and iv) a multi receptor tyrosine kinase inhibitor (mRTKI); b) administering a second treatment to the subject about one to four weeks after administration of the first treatment, the second treatment comprising a UGT inhibitor such as magnolol or a derivative thereof; c) administering a third treatment to the subject at least one week after administration of the second treatment, the third treatment comprising a 4-anilinoquinazoline, optionally in combination with an allosteric inhibitor, wherein the 4-anilinoquinazoline is selected from: v) a MARK inhibitor; vii) a P13K-Akt inhibitor; and viii) a mRTKI.

[0043] The allosteric inhibitor of the first and third treatments may be the same or different and may be GNF-2, GNF-5 or GN012380.

[0044] The second treatment may be administered to the subject a second time, one to three weeks after the first administration of the second treatment. The second treatment may be administered about two weeks after the first treatment and again about two weeks later; and the third treatment may be administered about two weeks thereafter.

[0045] The Bcr-Abl inhibitor of the first treatment may be a second generation Bcr-Abl inhibitor selected from dasatinib, bosutinib, axitinib, nilotinib or a derivative thereof, more particularly dasatinib or bosutinib. Alternatively, the Bcr-Abl inhibitor may be a third generation Bcr-Abl inhibitor such as ponatinib or rociletinib, or a derivative thereof, more particularly ponatinib.

[0046] The MAPK pathway inhibitor of the first treatment may be vemurafenib, dabrafenib, trametinib, cobimetinib or binimetinib.

[0047] The PI3K-Akt inhibitor of the first treatment may be idelalisib or ibrutinib.

[0048] The mRTKI of the first treatment may be lapatinib, sunitinib, sorafenib or pazopanib.

[0049] According to a further aspect of the invention, there is provided a method of treating keloid disease (KD) in a human subject, the method comprising the steps of: a) administering a first treatment to the subject, the first treatment comprising a 4- anilinoquinazoline and optionally also an allosteric inhibitor, wherein the 4- anilinoquinazoline is a pan-Bcr-Abl inhibitor selected from the group consisting of: i) a second generation Bcr-Abl inhibitor such as dasatinib, bosutinib, axitinib, nilotinib or a derivative thereof; ii) a third generation Bcr-Abl inhibitor such as ponatinib or rociletinib; iii) a MAPK pathway inhibitor such as vemurafenib, dabrafenib, trametinib, cobimetinib or binimetinib; iv) a PI3K-Akt inhibitor such as idelalisib or ibrutinib; or v) a mRTKI, such as lapatinib, sunitinib, sorafenib, or pazopanib; b) administering a second treatment to the subject one to four weeks after administration of the first treatment, the second treatment comprising a UGT inhibitor such as magnolol or a derivative thereof, optionally in combination with a 4-anilinoquinazoline such as axitinib, nilotinib, dasatinib or bosutinib; c) administering a third treatment to the subject at least one week after administration of the second treatment, the third treatment comprising a 4-anilinoquinazoline, wherein the 4- anilinoquinazoline is selected from the group consisting of: vi) a MAPK inhibitor selected from vemurafenib, dabrafenib, trametinib, cobimetinib, trametinib and binimetinib; and / or vii) a P13K-Akt inhibitor such as idelalisib or ibrutinib; or viii) a mRTKI inhibitor such as lapatinib, sunitinib, sorafenib, or pazopanib.

[0050] According to a further aspect of the invention, there is provided the use of a compound selected from the group consisting of dasatinib, bosutinib, axitinib, nilotinib, ponatinib, idelalisib, ibrutinib, vemurafenib, dabrafenib, trametinib, cobimetinib, binimetinib, lapatinib, sunitinib, sorafenib, pazopanib, magnolol, GNF-2, GNF-5 and ON012380, or derivatives thereof, in the manufacture of a medicament for use in treating keloid disease according to the method described above.

[0051] BRIEF DESCRIPTION OF THE FIGURES

[0052] Figure 1 : Schematic overview of some of the keloid disease (KD) treatment protocols described herein.

[0053] Figure 2: Two-way t-distributed stochastic neighbor embedding (t-SNE) shows groupings based on mRNA transcription data for cases (KD and FKN) and controls (NS and N) during wound healing modeled in vivo. The different groupings are assigned different shapes for sequential time points (day 0, 14 and 28). Each data point in the plot represents an individual’s mRNA profile. The N and NS data shows differential mRNA expression pattern during wound healing, as day 0 samples had a distinct cluster whereas day 14 and 28 formed a different cluster which had a higher overlap. On the contrary, KD and FKN formed clusters that showed static expression patterns during wound healing.

[0054] Figure 3: Kinase enrichment analysis scores from the X2K analysis model showing the top 20 protein kinases with the highest-ranking kinase as MAPK1 kinase. MAPK1 , Mitogen-Activated Protein Kinase 1 ; MAPK14, Mitogen-Activated Protein Kinase 14, CDK4, Cyclin Dependent Kinase 4, CDK1 , Cyclin Dependent Kinase 1 ; CSNK2A1 , Casein Kinase 2 Alpha 1 ; HIPK2, Homeodomain Interacting Protein Kinase 2; AKT1 , AKT Serine / Threonine Kinase 1 , ERK2, Extracellular Signal- Regulated Kinase 2; GSK3B, Glycogen Synthase Kinase 3 Beta; MAPK3, Mitogen -Activated Protein Kinase 3; GSK3BETA, Glycogen Synthase Kinase 3 Beta; CDK2, Cyclin dependent kinase 2; ERK1 , Extracellular Signal-Regulated Kinase 1 ; CK2ALPHA, Casein Kinase 2 Alpha 1 ; PKBALPHA, Protein Kinase B alpha; MAPK8, Mitogen -Activated Protein Kinase 8; PRKCZ, Protein Kinase C Zeta; JNK1 , Mitogen-Activated Protein Kinase 8 Isoform JNK1 Alphal ; TGFBR2, Transforming Growth Factor Beta Receptor 2; IKKALPHA, l-Kappa-B Kinase-Alpha; DNAPK, DNA-Dependent Protein Kinase Catalytic Subunit. Figure 4: Illustration of the expression patterns for EPHA8 (A) and MAPK12 (B) during the process of wound healing (day 0, day 14 and day 28) for KD compared to N. Dotted lines represent KD, solid lines represent N, solid stars indicate significant upregulation and white (no-fill) stars represent significant downregulation. EPHA8 was significantly upregulated on days 0 and 28. There was a significantly upregulation of MAPK12 on day 0 and a significant downregulation on day 28.

[0055] Figure 5: Overview of the integrative proteomics and transcriptomics analysis performed in this study to select KD specific biomarkers during the wound healing process.

[0056] Figure 6: Confirmatory qPCR for ANGPTL2, MAPK12 and SERPINH1 from a cohort of 18 KD and 4 N participants. Normalised fold change expressions with error bars presented in standard error of the mean (SEM) showing significant upregulation in the KD group compared to N group for A) ANGPTL2, B) MAPK12 and C) SERPINH1. *P< 0.05.

[0057] Figure 7: Sum of count for the top 25 compounds (of 1 ,005) identified in this study for treating

[0058] KD. The shading shows the sum of mean Combined Score, showing the extent to which the compound is likely to reverse the keloid signature across cell lines. The marks are labelled by the instance that the compound appears in the list of drugs across the cell lines.

[0059] Figure 8: Bar graph showing results of cytotoxicity assays for GNF-2, nilotinib, magnolol and bosutinib.

[0060] DETAILED DESCRIPTION

[0061] A method of treating a skin condition or disease, such as keloid disease (KD) or folliculitis keloidalis nuchae (FKN), is described herein. Compounds and combinations thereof for use in the method of treatment and a kit are also provided.

[0062] In this study, the Applicant used a temporal sequential biopsy approach on days 0, 14 and 28 in a cutaneous skin scarring model in three groups of human volunteer participants with KD, folliculitis keloidalis nuchae (FKN) and normal flat scar (NS) to elucidate KD pathogenesis. Normal skin (N) subjects were used as controls. RNA sequencing and proteomics were performed in all four groups. EPHA8 and MAPK12 were identified as candidate drivers for KD pathogenesis on day 0. Additionally, UDP-glucuronosyltransferases (UGTs) were identified as crucial drivers for KD development on days 14 and 28. Integrated transcriptomics and proteomics enrichment analyses highlighted seven dysregulated pathways, which suggest viral and cancer-like mechanisms. MAPK12 was one of the genes annotated to the common pathways. Two missense variants that had an effect at the protein level were found in two of the KD samples. Variant calling analysis found a missense variant in the IRS1 (where Ala changed to Ser in GSH-KD-009) and a missense variant in PDE3B (where Pro 23 changed to a Leu).

[0063] The treatment described herein includes administering to a patient suffering from KD one or more breakpoint cluster region-Abelson (Bcr-Abl) inhibitors, and optionally also one or more compounds which target EPHA8 and / or MAPK12 and / or one or more UGT inhibitors.

[0064] At least one of the compounds used in the treatment described herein is a breakpoint cluster region- Abelson (Bcr-Abl) inhibitor. Bcr-Abl inhibitors bind to the Bcr-Abl protein, a cytoplasmic fusion oncoprotein with constitutive tyrosine kinase activity and which is responsible for acute lymphoblastic and chronic myeloid leukemias. Binding of the Bcr-Abl inhibitor bind to the Bcr-Abl protein inhibits the activity of the protein, thereby preventing the growth and proliferation of leukemia cells. Allosteric inhibitors bind allosterically to Bcr-Abl mutations resulting from the EPHA8 reaction with ATP inhibitors. These molecules are non-ATP competitive and bind to the substrate binding region.

[0065] Suitable allosteric inhibitors of Bcr-Abl for use in this method include GNF-2, GNF-5 (the N- hydroxyethyl carboxamide analog of GNF-2) and ON012380. In one embodiment, the Bcr-Abl inhibitor is GNF-2. These allosteric Bcr-Abl inhibitors are not in the 4-anilinoquinazoline family.

[0066] GNF-2 GNF-5 ON012380

[0067] The allosteric Bcr-Abl inhibitor is typically administered to the patient at the time of starting treatment or at the time of surgery to remove a keloid (day 0). In the cases where surgery is performed, the allosteric inhibitor can be administered shortly prior to surgery, during surgery or shortly after surgery (typically within a few days of surgery). The allosteric inhibitor can be administered as a once-off single dose or can be administered in two or more doses at timed intervals, typically over a period of several weeks and more particularly, over a period of from 2 to 6 weeks.

[0068] The method of treatment can also include the step of surgically removing the keloid scar. In one embodiment, the allosteric Bcr-Abl inhibitor (e.g. GNF-2) is administered on its own, i.e. without any other drugs being administered. One or more 4-anilinoquinazolines can also be administered to the patient to treat the skin condition or disease. 4-anilinoquinazolines are a class of chemical compounds that belong to the quinazoline family. They are characterized by the presence of an aniline group (an amino group attached to a phenyl ring) at the 4-position of the quinazoline ring system. Compounds in this class include imatinib, dasatinib, bosutinib, axitinib, nilotinib, erlotinib, gefitinib, afatinib, dacomitinib, lapatininb, ponatinib, vemurafenib, dabrafenib, trametinib, cobimetinib, trametinib, binimetinib, sunitinib, sorafenib, pazopanib, bafetinib, rebastinib, tozasertib, danusertib, and TAK-285. The structures of some of these compounds are shown below. Nilotinib

[0069]

[0070] 4-anilinoquinazolines / type II tyrosine kinase inhibitors

[0071] 4-anilinoquinazolines were developed to be tyrosine kinase inhibitors (TKIs) and in particular type II TKIs. Some of these compounds have also been identified as Bcr-Abl inhibitors. Type II TKIs target receptor tyrosine kinases (RTKs) by binding to the kinase domain of the target protein in an inactive conformation. This binding mode distinguishes them from Type I inhibitors, which bind to the active conformation of the kinase domain. Type II TKIs typically bind to the ATP-binding site of the kinase domain. They take advantage of a specific inactive conformation where the "gatekeeper" residue, which controls access to the ATP-binding pocket, is displaced, allowing the inhibitor to bind. Upon binding, Type II inhibitors stabilize the inactive conformation of the kinase domain, locking it in an inactive state. This inhibits the enzymatic activity of the kinase and prevents downstream signaling. Type II TKIs often exhibit a higher degree of selectivity compared to Type I inhibitors. By targeting the inactive conformation, they can specifically bind to kinases with unique structural features in the inactive state.

[0072] The 4-anilinoquinazolines can be grouped within three generations of Bcr-Abl inhibitors, each generation representing a different class of drugs with varying levels of potency and specificity. The first-generation inhibitors include imatinib (Gleevec®). Imatinib was the first Bcr-Abl inhibitor approved by the FDA in 2001 for the treatment of chronic myeloid leukemia (CML). Second- generation inhibitors were developed to overcome resistance that some CML patients developed to first-generation inhibitors. These inhibitors are more potent and have broader activity against certain mutations that can arise in the BCR-ABL fusion gene, making the Bcr-Abl less responsive to imatinib. Examples of second-generation inhibitors include dasatinib (Sprycel®), nilotinib (Tasigna®), bosutinib and axitinib. The third-generation inhibitors were developed to address the resistance that some patients developed to both first- and second-generation inhibitors, and in particular to target the T315I mutation that emerged in the Bcr-Abl gene. Ponatinib (Iclusig®) is the most well-known third-generation Bcr-Abl inhibitor. It inhibits both the native and mutant Bcr-Abl protein, including the T315I mutation, and is hence sometimes referred to as a “pan- Bcr-Abl inhibitor”. Other examples of third generation Bcr-Abl inhibitors are rociletinib, idelalisib, ibrutinib, vemurafenib, dabrafenib, trametinib, cobimetinib, binimetinib, lapatinib, sunitinib, sorafenib and pazopanib.

[0073] The allosteric Bcr-Abl inhibitor can be administered in combination with a 4-anilinoquinazoline, such as dasatinib, nilotinib, bosutinib, axitinib, ponatinib, rociletinib, vemurafenib, dabrafenib, trametinib, cobimetinib, binimetinib, lapatinib, sunitinib, sorafenib or pazopanib, or a derivative thereof.

[0074] In one embodiment, the 4-anilinoquinazoline is a second generation Bcr-Abl inhibitor, i.e. dasatinib, nilotinib, bosutinib or axitinib, or a derivative or pharmaceutically acceptable salt thereof. More particularly, the 4-anilinoquinazoline is nilotinib. This treatment option is typically selected for a keloid scar which is not known or suspected to have an aggressive or recurring phenotype. Indicators of when a keloid scar may be determined to have an aggressive or recurring phenotype include, but are not limited to: multiple keloid scars being present on the body of the subject; multiple episodes of recurrence; where the scar rapidly increases in size, e.g. where there are clinically perceivable changes in size over a time period (e.g. of about 6 weeks); where the scar continues to grow beyond the boundary of the original lesion; where the scar is disfiguring; where the scar has not responded to previous treatment. In another embodiment, in particular for cases where the keloid scar is known or suspected to be aggressive or recurring, the 4-anilinoquinazoline can be selected from the following: a third generation Bcr-Abl inhibitor (pan- Bcr-Abl inhibitor), such as ponatinib or rociletinib, because these inhibitors are effective against a wider range of mutations than second generation inhibitors; a MAPK pathway inhibitor such as vemurafenib, dabrafenib, trametinib, cobimetinib or binimetinib; a PI3K-Akt inhibitor such as idelalisib or ibrutinib; or a multi receptor tyrosine kinase inhibitor (mRTKI), such as lapatinib, sunitinib, sorafenib, or pazopanib. As used herein, a mRTKI is a compound that can inhibit multiple receptor tyrosine kinases.

[0075] As mentioned above, a superfamily of enzymes called UDP-glucuronosyltransferases (UGTs) were identified during early wound healing (days 14 and 28) of the human model. These enzymes are known for breaking down drugs and xenobiotics. Optionally, the method of treatment can therefore also comprise administering to the subject a compound or composition that has an inhibitory effect on UGTs, including compounds that target phase II metabolic reactions driven by UGTs (hereinafter referred to as “the second treatment”). Magnolol, and derivatives thereof, are examples of UGT inhibitors.

[0076] Chemical structure of magnolol

[0077] The UGT inhibitor can be administered to the patient at least once from about 7 days post-surgery, such as at least 7 days, at least 10 days, at least 12 days, at least 14 days, at least 16 days, at least 3 weeks, at least 4 weeks post-surgery and so on. The UGT inhibitor can be administered in combination with axitinib, nilotinib, dasatinib or bosutinib.

[0078] In cases where the second treatment is administered more than once (i.e. repeated), the second treatment can be administered to the subject from about one to three weeks after it was previously administered, e.g. at about 3 or 4 weeks post-surgery (such as on about day 28) or even later.

[0079] KD development can continue beyond 6 weeks post-surgery and in this study, it was found that multiple MAPK pathway events were significantly enriched in some KD scars that were more than 1 year old. Optionally, a further treatment can therefore be administered to the patient for MAPK signal transduction at about weeks 5 or 6 post-surgery, or even later (hereinafter referred to as “the third treatment”). In one embodiment, the third treatment incudes a 4-anilinoquinazoline which is a MAPK inhibitor (such as vemurafenib, dabrafenib, cobimetinib, trametinib or binimetinib). In another embodiment, the third treatment incudes a 4-anilinoquinazoline which is a PI3K-Akt inhibitor (such as idelalisib or ibrutinib), as integrated proteogenomic analysis showed the PI3K-Akt signaling pathway to be commonly dysregulated / enriched. In another embodiment, the third treatment includes a combination of the MAPK inhibitor and the PI3K-Akt inhibitor. Alternatively, for aggressive or recurring KD scars, the third treatment can include a 4-anilinoquinazoline which is a mRTKI, since drug resistance to these molecules is rare. Suitable mRTKIs include sorafenib, sunitinib, lapatinib, pazopanib, axitib, vandetanib, lenvatinib, regorafenib and cabozantinib. The mRTKI is typically administered in combination with an allosteric inhibitor such as GNF-2, GNF-5 or ON012380.

[0080] In one embodiment, the treatment comprises the following two-step protocol:

[0081] 1 ) administering a first treatment to the patient, the first treatment comprising an allosteric Bcr-Abl inhibitor, optionally in combination with a 4-anilinoquinazoline; and

[0082] 2) administering a second treatment to the subject one to four weeks after administration of the first treatment, the second treatment comprising a UGT inhibitor such as magnolol or a derivative thereof, optionally in combination with a 4-anilinoquinazoline such as axitinib, nilotinib, dasatinib or bosutinib.

[0083] In another embodiment, the treatment comprises the following three-step protocol:

[0084] 1 ) administering a first treatment to the patient, the first treatment comprising an allosteric Bcr-Abl inhibitor, optionally in combination with a 4-anilinoquinazoline;

[0085] 2) administering a second treatment to the subject one to four weeks after administration of the first treatment, the second treatment comprising a UGT inhibitor such as magnolol or a derivative thereof, optionally in combination with a 4-anilinoquinazoline such as axitinib, nilotinib, dasatinib or bosutinib; and

[0086] 3) administering a third treatment to the subject at least one week after administration of the second treatment, the third treatment comprising a 4-anilinoquinazoline, wherein the 4-anilinoquinazoline is selected from the group consisting of: a MAPK inhibitor selected from vemurafenib, dabrafenib, trametinib, cobimetinib, trametinib and binimetinib; and / or a P13K-Akt inhibitor such as idelalisib or ibrutinib; or a MRTKI inhibitor such as lapatinib, sunitinib, sorafenib or pazopanib, optionally in combination with an allosteric Bcr-Abl inhibitor.

[0087] Figure 1 shows examples of multi-step treatment protocols for treating keloid disease. The compounds of each treatment step can be combined in a single composition or they can be administered separately. The compounds can be topically, transdermally, orally or intralesionally administered, e.g. by intralesional injection or by transdermal patch. Nanoparticles may be used to deliver the compounds or combination of compounds for each treatment. For example, a method of using nanoparticles to deliver a combination of drugs is described in Detappe et al.

[0018] , the contents of which are incorporated herein.

[0088] The method of treatment can thus also include a step of testing a sample of the keloid for one or more of the following biomarkers: EPHA8, MAPK12, DYRK2, MAPK11 , PRKD1 , PRKCQ, PAK1 , PRKG1 , MAP4K4, LYN, AKT3, IRAKI , JAK3, PRKACB, NEK6, BMPR1 B, FGFR1 , ABL2, IGF1 R, CAMK2A, UGT1A6, UGT1A7, SERPINH1 , ANGPTL2, UGT1A1 , UGT2A1 , UGT1 A10, UGT1 A3, UGT 1 A9, and UGT 1 A8. In one embodiment, the sample is tested for EPHA8 and / or MAPK12 before the subject is treated.

[0089] The expression levels of one or more of the following genes can also be determined: OMD, FAP, RCN1 , TNC, COL12A1 , CALD1 , THBS2 / TSP2, COL6A2, and PDIA5.

[0090] The presence of these markers or the overexpression of these genes (relative to a sample from an adjacent normal skin region (i.e. a region which does not have any signs of a keloid scar or any other scar tissue)) can be used to determine that the keloid scar has an aggressive phenotype.

[0091] A kit for treating KD is also provided. The kit can include any two or more of the following: a Bcr-Abl inhibitor which is an allosteric inhibitor, e.g. GNF-2, GNF-5 or ON012380; a 4-anilinoquinazoline selected from axitinib, nilotinib, dasatinib, bosutinib, ponatinib, vemurafenib, dabrafenib, trametinib, cobimetinib, trametinib, binimetinib, lapatinib, sunitinib, sorafenib, and pazopanib, or any combination thereof; a UGT inhibitor such as magnolol or a derivative thereof; and / or instructions for treating a patient with KD according to the method described herein.

[0092] The use of a compound selected from the group consisting of dasatinib, bosutinib, axitinib, nilotinib, ponatinib, idelalisib, ibrutinib, vemurafenib, dabrafenib, trametinib, cobimetinib, binimetinib, lapatinib, sunitinib, sorafenib, pazopanib, magnolol, GNF-2, GNF-5 and ON012380, or derivatives thereof, in the manufacture of a medicament for use in treating keloid disease according to the method described herein is also provided.

[0093] The invention will now be described in more detail by way of the following non-limiting examples. Examples

[0094] Methods

[0095] Participant recruitment, sample collection of sequential biopsies and sample processing

[0096] Ethics approval for the study was obtained from the University of Cape Town’s Human Research Ethics Committee (HREC REF:375 / 2017). The study was carried out following the ethical guidelines outlined in the declaration of Helsinki.

[0097] RNA sequencing (RNA-seq) and proteomics using liquid chromatography-coupled tandem mass spectrometry (LC-MS / MS) was performed using sequential biopsies in a cutaneous scarring model in humans (days 0, 14 and 28). The cohort of participants were cases: KD and folliculitis keloidalis nuchae (FKN) and controls: normal scar (NS) and normal skin (N) with an average age of 36,2 years. The KD and FKN patients had not undergone steroid injection for at least one year. The study included 15 participants from which 71 RNA-seq and 56 proteomics specimens were used after quality control procedures.

[0098] Skin biopsies were collected by Dermatology registrars from the Hair and Skin Clinic. On day 0, biopsy sites were marked for KD participants (margin and intralesional / centre of lesion); for FKN two biopsy sites were selected (margin of lesion and normal skin from inner arm) and one biopsy site was selected for NS and N. Four punch biopsies (4 mm) were taken from each site after numbing the area with a local anesthetic agent. The four biopsies were processed as follows: one was stored in AllProtect tissue reagent (Qiagen), the second biopsy was snap frozen in liquid nitrogen, for both RNA and protein extractions, and the third biopsy was stored in 10% buffered formalin solution (Sigma-Aldrich, Poole, UK) for histological analysis. The fourth and intralesional (in the case of KD) biopsies were collected in Dullbecco’s modified essential medium supplemented with 10% fetal calf serum (PAA Laboratories, Colber, Germany), and these were used for tissue culture (for another study). The first set of biopsies were re-biopsied on day 14 (from each lesion site of two weeks old healing tissue). The first biopsy was stored in AllProtect tissue reagent (Qiagen) for both RNA and protein extractions, whereas the second biopsy was stored in 10% buffered formalin solution (Sigma- Aldrich, Poole, UK) for histological analysis and immunohistochemistry. On day 28, the last set of biopsies from day 0 (from each lesional site of four weeks old healing tissue) were re-biopsied. The two biopsies were processed in a similar manner as those taken on day 14.

[0099] RNA and protein extractions

[0100] Tissue biopsies (4 mm) were collected from a -80°C freezer and kept in liquid nitrogen for processing. The tissue was diced with a cold sterile scalpel and blade, and immediately transferred into cold 2 mL screw cap tube containing 5 ceramic beads (Omni International, Kennesaw GA, USA) and snap frozen in liquid nitrogen. Tissue lysis was carried out on the Omni BeadRuptor24 (Omni International, Kennesaw GA, USA) for 3 cycles of 20 sec at 5 m / s at 0 - 4°C with snap freezing between cycles. After the third cycle, 500 pL of TriZol® reagent (Thermo Fisher Scientific) was added to the tube and lysis was carried out for 30 sec at 5 m / s at 0 - 4°C. The homogenate was transferred to a pre-chilled 1 .5 mL Eppendorf tube, and an additional 500 pL of TriZol was added to the 2 mL screw cap tube to further lyse residual tissue as mentioned above. The lysate was added to the initial lysate in a 1.5 mL Eppendorf tube to make up 1 mL and centrifuged for 2 min at 15 000 xg to precipitate debris. The supernatant was transferred to a new 1.5 mL Eppendorf tube containing 200 pL of cold chloroform for RNA extraction.

[0101] The tissue lysate and chloroform mixture was inverted several times (10 - 15 times) and stored at RT for 2 - 3 min followed by centrifugation at 18 000 xg for 15 min. The clear aqueous layer was carefully removed with a pipette and transferred to a new tube containing 10 pL of 3 M sodium acetate and 400 pL of 100% ethanol. The interphase was set aside for protein extraction. These were inverted several times and allowed to stand for 1 min on ice to precipitate RNA. The AHPrep DNA / RNA / Protein mini kit (Qiagen) protocol was followed with minor modifications for RNA extraction. Briefly, 700 pL mixture was transferred to the pink RNeasy spin column from the kit and centrifuged at 14 000 xg for 15 sec. The flow-through was discarded and 700 pL of cold Buffer RW1 was added to the RNeasy column and centrifuged at 13 000 xg for 30 sec. Subsequently, 500 pL RPE buffer was added to the column and centrifuged at 13 000 xg for 30 sec and the flow-through was discarded. A 500 pL volume of cold 70% ethanol was added to the RNA column and centrifuged at 14 000 xg for 2.5 min. The RNeasy column was placed on new 1.5 mL tube and 30 pL of cold RNase free water added directly on the column and allowed to stand for 30 sec to 1 min at RT. RNA was eluted by centrifugation at 14 000 xg for 1 min. RNA was aliquoted into other tubes and subsequently stored at -80°C for RNA sequencing.

[0102] The precipitated debris resulting from lysis was mixed with the white protein interphase that was set aside for protein extraction. A 1 volume (600 - 1000 pL) of 100% ethanol was added to these and mixed vigorously and incubated at RT for 10 min followed by centrifugation at 20 000 xg for 10 min. The supernatant was carefully discarded and 500 pL of 75% ethanol was added to the pellet and centrifuged at 20 000 xg for 1 min. The supernatant was removed carefully using a pipette. The pellet was dried for 5 - 10 min at RT and eventually re-suspended in 50-100 pL of a 1 :1 ratio of 8M urea and 1 % SDS solution with subsequent vortexing. These were incubated for 5 min at 95°C to completely dissolve and denature the protein. Vortexing was done several times to aid with resuspension. Subsequently, centrifugation was carried out at 15 000 xg for 30 sec to precipitate the residual insoluble material. The protein was quantified and stored at -80°C. RNA sequencing

[0103] The RNA sequencing (RNA-seq) library preparation method of choice was the strand-specific and paired-end to facilitate a thorough RNA-seq analysis [1 , 2], The nature of the hard, fibrotic tissue (KD and FKN) yielded degraded RNA whilst NS and N samples yielded variable but better quality RNA. As a result, a ribo-depletion method was selected for library construction to avoid 3’ bias or library construction impairment as well as to have an holistic view of the transcriptome in the study cohort. Library construction and RNA-seq was carried out as a service by BGI Global Genomics (www.bgi.com / global / ]. The extracted RNA from N, NS, KD and FKN was shipped on dry ice to BGI Global Genomics. Additional RNA quality was verified using a 2100 Bioanalyzer instrument (Agilent Technologies, California, USA). RNA-seq libraries were prepared using a MGIEasy rRNA depletion kit (Shenzhen, China) and MGIEasy RNA directional library prep kit (Shenzhen, China) according to the manufacturer’s instructions.

[0104] Proteomics

[0105] Proteomics was carried out as a service by the Centre for Proteomic & Genomic Research (CPGR) (www.cpgr.org.za / ) in Cape Town, South Africa. Protein concentrations were quantified using a bicinchoninic acid (BCA) assay [3]. In preparation for the HILIC magnetic bead workflow, HILIC beads (ReSyn Biosciences, HLC010) were aliquoted into a new tube and the shipping solution removed. Beads were then washed with 250 pL wash buffer (15% Acetonitrile (ACN), 100 mM Ammonium acetate (Sigma 14267) pH 4.5) for one minute. This was repeated once. The beads were then resuspended in loading buffer (30% ACN, 200 mM Ammonium acetate pH 4.5). A total of 50 pg of protein (or the entire sample) from each tube was transferred to a 96-well protein LoBind plate (Merck, 0030504.100). Protein was reduced with tris (2-carboxyethyl) phosphine (TCEP; Sigma 646547) which was added to a final concentration of 10 mM TCEP and incubated at 30“C for two hours (urea containing samples) or 60 “C for one hour (all other samples). Samples were cooled to RT and then alkylated with methylmethanethiosulphonate (MMTS; Sigma 208795) which was added to a final concentration of 10mM MMTS and incubated at RT for 15 minutes. HILIC magnetic beads were added at an equal volume to that of the sample and a ratio of 5:1 total protein. The plate was then incubated at RT on the shaker at 900 RPM for 30 minutes for binding of protein to beads. After binding, the beads were washed four times with 500 pL of 95% ACN for one minute. For digestion, Trypsin (Promega PRV51 11 ), was made up in 50mM Triethylammonium bicarbonate buffer (TEAB) and added at a ratio of 1 :10 total protein. The plate was then incubated at 37“C on the shaker for 4 hr. After digestion, the plate was placed on a magnet to separate the beads from the supernatant. The supernatant containing peptides was removed and dried down. Samples were then resuspended in LC loading buffer: 0.1 % FA, 2.5% ACN. LC-MS analysis was conducted on a Q-Exactive Hyvrud Quadrupole-Orbitrap mass spectrometer (Thermo Fisher Scientific, USA) coupled to a Dionex Ultimate RS3000 nano-UPLC system. Data was acquired using: Xcalibur v4.1 .31 .9, Chromeleon v6.8 (SR13), Orbitrap MS v2.9 (build 2926) and Thermo Foundations 3.1 (SP4). Peptides were dissolved in 0.1 % Formic Acid (FA; Sigma 56302), 2% Acetonitrile (ACN; Burdick & Jackson BJLC015CS) and loaded on a C18 trap column (PepMapl OO, 300 pm x 5 mm x 5 pm). The solvent system employed was solvent A: LC water (Burdick and Jackson BJLC365); 0.1% FA and solvent B: ACN, 0.1 % FA. Samples were trapped onto the pre-analytic column at 2% solvent B and washed for 3 minutes before the valve was switched and peptides eluted onto the analytical column. Chromatographic separation was performed with a Waters nanoEase (Zenfit) M / Z Peptide CSH C18 analytic column (186008810, 75 pm x 25 cm x 1 .7 pm). The solvent system employed was solvent A: LC water (Burdick and Jackson BJLC365), 0.1% FA and solvent B: ACN, 0.1% FA. The multi-step gradient for peptide separation was generated at 300 nL / min as follows: time change 1 17 min, gradient change: 5 - 40% solvent B, time change 5 min, gradient change 40 - 80% solvent B. The gradient was then held at 80% solvent B for 10 min before returning it to 5% solvent B and conditioning the column for 15 min.

[0106] Proxeon stainless steel emitters (Thermo Fisher TFES523) were used for all spectral data acquisition. Mass spectrometry was operated in positive ion mode with a capillary temperature of 320°C with the applied electrospray voltage of 1 .95 kV.

[0107] Computational analyses

[0108] Prediction of upstream regulator kinases

[0109] The Expression2Kinases (X2K) analysis model reported by Chen and co-authors [4] was employed to predict the upstream regulatory kinases that likely give rise to the keloid phenotype at the cellular level. The differentially expressed gene lists from the pairwise analysis from were fed into the transcription factor inference module of X2K using the database ChlP-seq / ChipX Enrichment Analysis (ChEA) to predict transcription factors that are likely to be upstream regulators from the differentially expressed gene list. The ChEA database was used to compute enrichment for the overlap between the differentially expressed genes between keloids and normal skin and entries in the ChEA database using a combination of the Fisher’s exact test and the computed the deviation from the expected rank for random gene list input. In the next analysis, these transcription factors were linked to upstream regulatory mechanisms by generation of a transcription factor-intermediate protein-protein interaction (PPI) sub-network based on known interactions using the computational tool, Genes2Network (G2N) [5]. This sub-network has 24 036 nodes (proteins) with 389 959 edges (interactions) and is enriched in co-regulators, kinases and other transcription factors that have been experimentally verified to physically interact with the 10 transcription factors. Finally, the sub-network was analyzed for enriched targets of known protein kinases likely to phosphorylate proteins within the sub-network using the computational tool, Kinase Enrichment Analysis (KEA) module of Expression to Kinase [4, 6].

[0110] Identification of viable drug target using network robustness and tolerance analysis (Barabasi-Albert model)

[0111] The removal or inhibition of the most connected nodes within a network is likely to maximally disrupt signaling through a signaling network. Using the Barabasi-Albert model [7, 8], such nodes were identified within the “keloid interactome”. Here, an interaction network was constructed using the list of proteins that were found to be upregulated in keloids compared to the normal skin. The connectivity between the network nodes was established prior to knowledge of protein-protein interactions extracted from the University of California Santa Cruz (UCSC) Super pathway [101 525 protein-protein interactions], Kinase Enrichment Analysis (KEA) database [428 Kinases; 10 792 interactions], and ChIP Enrichment Analysis (ChEA) database [667 transcription factors; 464 967 interactions]. Each protein (node) within the network was removed from the network (simulating loss of function mutation or drug inhibition) one at time, while computing the network diameter at each subsequent step. The network diameter was applied as proxy for the network’s ability to propagate information (continue signaling) in the absence of the removed node. Here, the diameter defines the shortest distance between the two most distant nodes in the network. The most important proteins (hub kinases) disrupt signaling through the network which is increasing the network diameter to some large value up to infinity (illustrating complete breakdown of the network). Therefore, the proteins were ranked based on the calculated diameter. This analysis was performed in MATLAB 2018b.

[0112] Variant calling

[0113] For the variant calling analysis, the candidate biomarkers were identified in the KEGG pathways and any genes upstream and downstream were included. From this, a bed file was created and Freebayes was used for variant calling. The results were filtered using vcffilter and annotated with SnpEff. This information was placed into a network to visualise where mutations were found, genes that were highly connected and the potential relationships between mutated genes and genes that were affected downstream.

[0114] Computational identification of compounds to treat keloids

[0115] The differentially expressed mRNA transcripts were identified between KD and N using a negative binormal model with a local regression model [9]. A list of the significantly upregulated genes (p < 0.05; log 2-fold change > 2) and downregulated genes was utilized (p < 0.05; log 2-fold change < - 2) to query a library of integrated network-based cellular signatures (LINCS)

[0010] L1000 characteristic direction signatures search engine (L1000CDS2)

[0011] . Using L1000CDS2, the cosine similarity was computed between the transcription signature that represents the transformation of the normal cells to keloid cells and the transcriptional signature of cell lines after perturbation with ~ 1 ,300 compounds.

[0116] The cosine similarity (distance) can be used to identify the drugs that produce transcription changes similar to those that occur during the transformation of the normal cell to keloids cells and also drugs that may reverse the transcription signatures. Simply the sign of the cosine similarity (computed as shown below) can provide such information:

[0117] Quantitative polymerase chain reaction (qPCR) qPCR was carried out in 10 pL reactions using Applied Biosystems QuantiStudio 6 Flex Real-Time PCR system (ThermoFisher Scientific, Massachusetts, USA). The reaction constituted: 1 x of 2x QuantiTect SYBR Green, 0.3 pM forward primer and 0.3 pM reverse primer, 2 pL cDNA from validation samples and nuclease-free waster (Qiagen). Reactions were carried out in triplicate using 96-well plates (ThermoFisher Scientific). The cycling conditions were as follows: 95 °C for 15 min to initiate the reaction, followed by 40 cycles of 95 °C for 15 sec (denaturation), 59 °C for 30 sec (annealing) and 72 °C for 30 sec (elongation). A melt curve of a 95 °C stage followed by a 60 °C stage and final 95 °C stage was included. qPCR was performed in a 384-well plate on a QuantStudio 12K Flex PCR instrument (ThermoFisher Scientific, Massachusetts, USA). GAPDH was used as the endogenous control for all relative gene expressions.

[0118] Immunohistochemistry (IHC)

[0119] Immunohistochemistry (IHC) was performed in 10% formalin-fixed, paraffin-embedded KD, FKN, NS and N 4pm tissues. Tissue sections were incubated overnight at 60 °C. The glass slides were dewaxed: (3x 1 min in xylene, 3x 1 min in absolute ethanol, 2x1 min in 96% ethanol, 2x 1 min in 70% ethanol, and 1 min in water). The tissue sections were blocked for non-specific staining using 3% hydrogen peroxidase for 15 min. The tissue sections were heated in a microwave submersed in 1x citrate buffer pH 6 unmasking solution until boiling was initiated; followed by 7 min at a sub-boiling temperature (95°- 98° C). The slides were cooled on top of a bench for 20 min. The sections were circled with Pap pen (hydrophobic pen) and washed 3 times with wash buffer (PBS / T). Tissue sections were then incubated overnight with primary antibody UGT1 A6 (rabbit polyclonal, 1 :100; Invitrogen, US), UGT1A7 (rabbit polyclonal, 1 :100; Invitrogen, US) and 1.5 hrs with primary Phospho-EphA8 (rabbit polyclonal, 1 :100; Invitrogen, US), Phospho-p38 MAPK gamma / delta (rabbit polyclonal, 1 :100; Invitrogen, US). The sections were washed carefully 3 times with wash buffer PBS-T. The sections were incubated with a secondary antibody (Detection reagent) (Goat Anti- rabbit, peroxidase antibody, Dako) for 30 min in a humidity chamber. The sections were washed three times with wash buffer PBS / T. 3,3-Dimainobenzidine DAB chromogen concentrated (One drop of 1 ml DAB diluent [UGT1 A6 and UGT1A7]) or Peroxidase system (1 drop to 2 ml PBS [phosphor- EPHA8 and MAPK12]) mixed well before adding to the sections. The sections were then incubated for 6 min with DAB or peroxidase (in the dark), generally to provide an acceptable staining intensity. The slides then immersed in water and incubated for 5 min in Haematoxylin and washed for 5 min in running tap water. The tissue sections were dehydrated (the sections were incubated in 70% ethanol two times for 1 sec each, incubated in 95% ethanol two times for 1 sec each, incubated in 100 % ethanol three times for 1 sec each and incubated in xylene three times 1 second each). The sections were mounted with mounting medium and coverslips.

[0120] Cell revival and culture conditions

[0121] Keloid disease fibroblasts and normal skin (from healthy female breast reduction surgery) fibroblasts were revived from liquid nitrogen storage. Vials containing frozen fibroblasts were thawed in a 37°C water bath and the vials were decontaminated by spraying with 70% ethanol. The cell suspension was transferred to a 15 mL centrifuge tube containing 9 mL of complete Dulbecco’s Modified Eagle’s Medium (cDMEM; DMEM GIBCO, Grand Island, NY) supplemented with 10% heat-inactivated foetal bovine serum (FBS), 1 % pen / strep. Cells were pelleted down by centrifuging the cell suspension at 125 x g for 5 min. The supernatant was aspirated, and the cell pellet was resuspended in 4 mL of cDMEM and seeded in T75 flasks. Cells were incubated in a humidified incubator (37°C, 95% O2, 5% CO2) until settled. The media was changed every 3 to 4 days until cells were confluent. Cells were used for experiments in lower passage number (< 10).

[0122] Cytotoxicity assays / half-maximal inhibitory concentration (!C5Q)

[0123] To establish the half-maximal inhibitory concentration (IC50), cytotoxicity assays were performed using (MTT) assay

[0012] , Confluent cells were harvested by washing twice with phosphate buffered saline (PBS, pH 7.4, Thermo Fisher Scientific) followed by trypsinisation with 3 to 4 mL of 1 x trypsin— EDTA [0.05% trypsin, 0.02% (0.53 mM) EDTA; Invitrogen Life Technologies] for 2 to 3 min. The effect of trypsin-EDTA was stopped by adding 6 to 8 mL cDMEM. The cell suspension was collected in a 15 mL centrifuge tubes and pelleted by centrifugation at 125 x g for 5 min. The supernatant was aspirated off and the cell pellet was resuspended in 4 mL of cDMEM. Cells were stained with trypan blue solution and counted using an automated cell counter (ThermoFisher Scientific). The cells were seeded in quadruplicate in 96-well plates (8000 cells / 200 pL / well) for KD and N fibroblasts and incubated for 24 hrs.

[0124] Nilotinib (Product #: CDS023093), axitinib (Product #: PZ0193), bosutinib (Product #: PZ0192), GNF- 2 (Product #: G9420), magnolol (Product #: PHL89317) and dexamethasone (Product #: D4902) were dissolved in dimethyl sulfoxide (DMSO, Merck). After 24 hrs incubation in 96-well plates, cells were treated with varying concentrations of each compound in DMEM supplemented with 2% heat- inactivated foetal bovine serum (FBS), 1% pen / strep. A blank served as media only (DMEM+2%FBS+1 %P / S) and a negative control untreated cells (DMEM+2%FBS+1%P / S) and a DMSO equivalent by volume of the highest drug concentration. After 24 hrs treatment, an MTT assay was performed by adding 100 pL of a 0.5 mg / mL of 3-(4,5-dimethylthiazol-2-yl)-2,5- diphenyltetrazolium bromide (MTT) dissolved in DMEM FBS-free. Plates were covered with foil and incubated for 3 hrs in the humidified incubator. Subsequently, the MTT solution was aspirated and 100 pL absolute ethanol added to each well. Absorbance was measured in the Varioskan microplate reader (ThermoFisher Scientific) at 570 nm. Experiments were performed in triplicate. The IC50 was calculated using non-linear regression from log transformed dose-response curves using GraphPad Prism v 10 (www.graphpad.com).

[0125] Checkerboard assay / combination

[0126] To evaluate synergism between the compounds, a checkerboard assay

[0012] was performed by combining varying concentrations of GNF-2 (0.125, 0.5, 1 , 5, 10, and 20 pM) with nilotinib IC50 (1 .90 pM) and1 / 2 IC50 (0.95 pM), and vice versa for varying concentrations of nilotinib (0.125, 0.5, 1 , 5, 10, and 20pM) and GNF-2 IC50 (21.08 pM) and1 / 2 IC50 (10.54 pM). The same procedure was performed for combinations of bosutinib and magnolol. The interaction between the drugs was calculated using the fractional inhibitory concentration (FIC)

[0012] , calculated using the formula:

[0127] The FIC interpretation was as follows: FIC < 1.0 indicates synergism; 1.0 < FIC < 2.0 indicates an additive effect; and FIC > 2.0 indicates antagonism [12, 13].

[0128] Statistical analysis

[0129] For drug testing experiments, statistical analyses were performed in GraphPad Prism and a P-value < 0.05 represented a statistically significantly different result.

[0130] Results

[0131] Transcriptomic exploratory analysis

[0132] Unsupervised hierarchical clustering using the top 500 most variable mRNA expression levels from KD, FKN, NS and N was performed. The clustergram resolved four distinct clusters based on sequential time points, with N and NS day 0 samples mainly forming their cluster while day 14 and 28 samples also formed a distinct cluster. KD and FKN samples did not seem to resolve based on sequential time points. The mRNA transcription data further portrayed three cluster patterns resolved by conditions (FKN, KD and N) while NS clustered evenly within KD and N. The transcription data was further explored using a one-way principal component analysis (PCA) plot over the three time points (days 0, 14, and 28) to visualise gene expression during wound healing. Day 0 served as the baseline where each data point represented gene expression changes per individual during wound healing. The first principal component (PC-1 ) accounted for 88% of the variance in the mRNA transcription data. Day 0 N samples clustered together along the PC-1 while by day 14, the expression pattern showed a distinguishable shift. By day 28 the N samples had shown a further shift in differential expression variance and clustered further upwards along PC-1. The NS samples followed similar patterns over the healing process as N samples. KD and FKN samples did not exhibit variance over the wound healing process suggesting static expression patterns regardless of the time point during wound healing.

[0133] Finally, the t-distributed stochastic neighbour embedding (t-SNE) analysis to the KD, FKN, NS and N was performed on the mRNA data (not shown). Each data point represented a participant and were assigned different shapes for time point, gender, age (Figure 2). KD and FKN samples corroborated PCA observations and formed their respective distinct clusters without exhibiting distinct gene expression patterns for the sequential time points during the wound healing process. The day 0 N and NS samples formed a distinct cluster while the day 14 and 28 samples co-clustered farther from the day 0 samples suggesting different gene expression profile for the two time points during wound healing. Furthermore, the day 14 and 28 gene expression patterns could exhibit overlapping expression patterns during remodelling stages of wound healing, hence the coclustering. Notably, the day 14 and 28 NS samples clustered closer to the KD main cluster suggesting that previously injured skin may express similar gene expression profiles during the day 14 - 28 healing process.

[0134] Proteomics exploratory analysis

[0135] Protein expression patterns were compared using MaxQuant intensity-based absolute quantification (iBAQ) values. Hierarchical clustering was subsequently used for the most significantly abundantly expressed protein groups across KD, FKN, scar / NS, and skin / N (not shown). KD samples mainly formed a distinct cluster with smaller sub-clusters of the respective time points. Additionally, individual participant samples clustered together. The FKN samples mostly clustered together and did not cluster according to time points of the wound healing process. There was a bigger overlap in the skin / N and scar / NS samples cluster. However, skin / N samples portrayed subclusters based on the time points. A two-way PCA plot was constructed for all conditions where PC-1 and PC-2 explained 24% and 12.2% of the variance, respectively (not shown). The KD samples formed a main cluster along PC-1. FKN samples clustered along PC2 and were closer to the skin / N and scar / NS sample clusters. There was a bigger overlap between the skin / N and scar / NS samples suggesting similar protein expression patterns.

[0136] Transcriptomic analysis

[0137] Gene expression profiles for all six pairwise analyses (KD-vs-N, KD-vs-NS, KD-vs-FKN, FKN-vs-N, FKN-vs-NS and NS-vs-N) were compared across each time point (day 0, 14 and 28) and differentially expressed genes (DEGs) were identified. Venn diagrams for both upregulated and downregulated DEGs were constructed. KD-vs-N and KD-vs-NS followed a similar pattern during the wound healing process and the NS-vs-N healing pairwise analysis showed minimal variances during the healing process with major difference at the baseline (day 0). KD-vs-FKN showed more variances on day 14 for the upregulated DEGs.

[0138] Identification of the most plausible candidate biomarkers using the overlap between the Expression2Kinases (X2K) and the Barabasi-Albert models

[0139] Prediction of upstream regulator kinases using the X2K model

[0140] The X2K model was used to analyse the DEGs from KD-vs-N day 0 for genes that correlated across all samples to predict upstream regulatory kinases that are responsible for observed gene expression patterns observed in KD tissue. Overall, 245 protein kinases and their respective ranks and hypergeometric values were identified. Three of the Protein kinase C (PKC) that were ranked differently could not be distinguished, suggesting 243 protein kinases overall. The top 20 kinase enrichment analysis scores are illustrated in Figure 3. These were predominantly from the MAP / ERK cascade (MAPK1 / ERK2, MAPK14, MAPK3 / ERK1 , MAPK8). The top five biomarkers were MAPK1, MAPK14, CDK4, CDK1 and CSNK2A1.

[0141] Identification of viable drug target using the Barabasi-Albert model

[0142] The “keloid interactome” was interrogated using the Barabasi-Albert model to identify the most connected nodes within the network that would serve as potentially viable drug targets. 71 out of the 1399 proteins could increase the diameter of the network to an infinite number, while the rest yielded a network diameter of 6 or less. An infinite network diameter implies that the removal of that specific protein from the keloid signalling network would completely disrupt signalling between various component of the network, i.e. , there are no alternative routes to the downstream targets[7, 8]. The top five biomarkers were AKT3, CTNNB1, ARHGAP28, ARHGEF7 and ATF3.

[0143] Table 1 : List of protein kinases with an infinite network diameter identified as viable drug targets using the Barabasi-Albert model

[0144] To identify the most plausible biomarkers for KD pathogenesis, the common targets shared between the 243 protein kinases and 1399 viable drug targets identified using X2K analysis and the Barabasi- Albert model, respectively, were compared. This identified 20 protein kinases (Table 2).

[0145] Table 2: List of overlapping protein kinases identified by both the X2K and Barabasi-Albert model

[0146] EPHA8 showed a significant upregulation for KD-vs-N on day 0 (logFC = 3.14, AdjP = 8.62 x 104) and day 28 (logFC = 1.86, AdjP = 6.73 x 103) (Figure 4, Table 2). Although FKN-vs-N day 0 also portrayed a significant upregulation (logFC = 2.70, AdjP = 1.96 x 102), KD-vs-FKN day 14 showed significant EPHA8 upregulation in KD (logFC = 4.15, AdjP = 1.14 x 102). Overall, EPHA8 exhibited unique expression patterns specific to the KD phenotype suggesting that it can be a viable KD marker.

[0147] MAPK12 also illustrated unique expression patterns specific for the KD phenotype (Figure 4, Table 2). On day 0, MAPK12 exhibited significant upregulation in all the KD pairwise analyses; KD-vs-N, KD-vs-NS, and KD-vs-FKN. Interestingly, MAPK12 was significantly upregulated on day 14 for KD- vs-FKN (logFC = 3.33, AdjP = 6.34 x 1010) whereas there was a significant downregulation in the FKN relative to N and NS. On day 28, MAPK12 showed a significant downregulation for KD-vs-N (logFC = -0.59, AdjP = 1.57 x 102) whereas it showed a significant upregulation in KD relative to FKN (logFC 3.24, AdjP = 6.41 x 10'9). Therefore, MAPK12 showed unique expression patterns, particularly for the KD phenotype.

[0148] Enrichment analyses

[0149] Transcriptomics

[0150] Enrichr [14, 15] was used to query for enriched KEGG and Reactome pathways using a list of gene sets identified using the X2K and Barabasi-Albert models, and the significantly upregulated genes from the conditions (KD, FKN, NS and N), and their sequential biopsies. When the enrichment pathway terms were sorted based on the combined score (CS), it was found that from the top 10 enriched KEGG and Reactome pathways, MAPK related pathway events were common from the X2K enrichment analysis. Additionally, MAP kinase activation in TLR cascade and ErbB signaling pathways were the top enriched terms for Reactome and KEGG, respectively. From the Barabasi- Albert model, ECM and collagen related events were found to be common in the top 10 significantly enriched KEGG and Reactome pathways. ECM-receptor interaction (CS = 110.51) and extracellular matrix organization (CS = 245.60) were the top enriched terms for the KEGG and Reactome pathway terms respectively. EPHA8 and other Eph receptors were annotated to the axon guidance, EPHA- mediated growth cone collapse, and Eph-ephrin mediated repulsion of cells pathway terms and Eph- ephrin signaling was annotated to either KEGG or Reactome pathway terms. Both EPHA8 and MAPK12 were annotated to the developmental biology term. MAPK12 was annotated to several KEGG and Reactome pathway terms.

[0151] For KD-vs-dO analyses, several previously reported KEGG and Reactome pathway terms were related to ECM, collagen, inflammatory events, and arthritis. The top enriched terms were ECM- receptor interaction and extracellular matrix organization pathway terms under KEGG and Reactome pathways respectively. Previously reported KD pathway terms such as PI3K-Akt signaling, IL-17 signaling pathway, focal adhesion, and rheumatoid arthritis were found. Some KD unreported pathways included relaxin signalling and AGE-RAGE signaling pathway in diabetic complications for which MAPK12 was the underlying gene in both KEGG pathway terms. Several ECM, viral mechanisms, MARK related events, relaxin signaling, cytokine receptor interactions, and inflammatory signaling mechanisms were common for X2K and Barabasi-Albert models and KD-vs- dO enrichment analysis.

[0152] For NS-vs-N-dO, the significantly enriched top KEGG pathway was protein digestion and absorption. Although this pathway overlapped with the KD-vs-N-dO enriched pathway term which included 15 collagens and the CPB1 gene, NS-vs-N-dO had 10 collagen genes. For FKN-vs-N-dO, the KEGG enriched top terms was rheumatoid arthritis. For Reactome terms, the top enriched term for NS-vs- N-dO, KD-vs-N-dO and FKN-vs-N-dO were different.

[0153] Glucuronidation and ascorbate and aldarate metabolism were the top enriched terms for Reactome and KEGG pathway terms for KD-vs-N on days 14 and 28. The predominant gene drivers for these pathways were UDP-glucuronosyltransferases (UGTs). Glucuronidation and interleukin-19, 20, 22, 24 Reactome pathway terms were significantly enriched for FKN-vs-N on days 14 and 28 respectively. For KEGG pathways, however, ascorbate and aldarate metabolism and pentose and glucuronate interconversions were significantly enriched on days 14 and 28 respectively. KD-vs-N and FKN-vs-N on days 14 and 28 predominantly had UGTs annotated to their enriched pathway terms, except for interleukin-19, 20, 22, 24. NS-vs-N-d14 did not show any significantly enriched pathways, whereas on day 28 the top enriched pathways were lnterleukin-1 processing and rheumatoid arthritis for the Reactome and KEGG pathways respectively. Proteomics

[0154] Enrichment analysis was performed on gene-sets from pairwise analyses between conditions (KD, FKN, NS, and N) and their respective time points (days 0, 14 and 28). The globally perturbed protein expression values in the database (KEGG or REACTOME pathway terms) were compared using Generally Applicable Gene-set Enrichment (GAGE)

[0016] and were visualised on the KEGG pathways using the R pathview package

[0017] , The pipeline for the analysis is available on GitHub (mqmetaproteomics v2.5) (github.com / uct-cbio / proteomics- pipelines / tree / master / singularity / mqmetaproteomics). For all KD pairwise analyses, there was a significant perturbation for the group of proteins annotated to seven common KEGG pathways (Tables 3.1 , 3.3 and 3.3). Interestingly, the relaxin signalling pathway was one of the top two ranked pathways for all KD group analyses, whereas the mTOR signalling pathway was the top-ranked pathway in KD relative to NS and FKN comparisons. For the KD-vs-N group, the top-ranked pathways were the relaxin signalling pathway and the PI3K-Akt signalling pathway. Within the relaxin signalling pathway, most KD samples exhibited upregulation of CREB1 , SRC and SHC1 during the wound healing process (days 0, 14 and 28). In the PI3K-Akt signalling pathway, proteins that exhibited continuous upregulation during wound healing were RHEB, EIF4B, RPS6, CDC37, CDK4 and CREB1 . Some of these dysregulated proteins are known oncogenes.

[0155] Significant upregulation was identified in the expression values of the group of proteins annotated to 22 Reactome pathway for the KD-vs-N groups. For KD-vs-N, the top-ranking Reactome pathways were selenocysteine synthesis and viral mRNA translation. These pathways were found across all KD pairwise analyses.

[0156] Table 3.1 : Enriched KEGG pathway terms for dysregulated proteins in the KD vs Skin group for combined time points (days 0, 14 and 28). Table 3.2: List of significantly perturbed KEGG pathway terms for the KD vs Scar pairwise analysis for all time points (days 0, 14 and 28).

[0157] Table 3.3: List of significantly perturbed KEGG pathway terms for the KD vs FKN pairwise analysis for all time points (days 0, 14 and 28).

[0158] Integrated proteomics and transcriptomics (“Proteo-transcriptomics”) - Functional analyses

[0159] An overview for “Proteo-transcriptomics” analysis is summarised in Figure 5. Enrichr and GAGE were used for the enrichment analyses for transcriptomics and proteomics, respectively. The seven significantly enriched KEGG pathway terms from proteomic analyses were also found significantly enriched in the transcriptomic analyses (Table 4). Additionally, Table 4 shows the respective transcriptomics analyses terms which overlapped with the proteomics analyses. The enriched pathway terms from the Barabasi-Albert model had 5 common pathway terms relaxin signalling, PI3K-Akt signalling pathway, human cytomegalovirus infection, human papillomavirus infection, and human immunodeficiency virus 1 infection), followed by the X2K model which shared 4 common pathway terms (highlighted). MAPK12 was one of the genes commonly found underlying the enriched transcriptomics pathway terms that overlapped with proteomics enrichment analyses. Table 4: Integrated proteomics and transcriptomics (proteo-transcriptomics) enrichment analyses overlap. The list of significantly dysregulated pathway terms identified in proteomics using the GAGE algorithm for KD vs Skin which overlapped with the pathway terms enriched in transcriptomics data. The 2ndcolumn represents the proteomics analyses and the 3rd-7thcolumns represents the different transcriptomics analyses which had common terms with the proteomics enrichment analyses. The highlighted blocks represent specific analyses that were enriched for specific pathway terms.

[0160] The data was subsequently interrogated by comparing both candidate biomarkers identified using the Barabasi-Albert model (1399 viable drug targets) and the X2K model (243 upstream regulators) with the overall identified proteins. This resulted in 40 mRNA / protein targets. Upon further interrogation, SERPINH1 and ANGPTL2 showed correlating direction of expression at least at one time point at both mRNA and protein level in the KD phenotype compared to N and FKN.

[0161] Differential gene expression of selected KD specific biomarkers using qPCR and immunohistochemistry qPCR

[0162] A new cohort of 22 participants was recruited for collection of samples to use for validation using qPCR from the four identified KD “proteo-transcriptomic” analyses biomarkers (EPHA8, MAPK12, SERPINH1 and ANGPTL2). The participants included 4 normal breast controls and 18 KD participants with an average age of 31 .82 years. The expressions of the candidate biomarkers were normalised to the endogenous control (GAPDH) and the expression represented as mean ± standard error of the mean (SEM). The qPCR analysis showed a significant upregulation for ANGPTL2, MAPK12 and SERPINH1 in the KD group compared to the N group (Figure 6). EPHA8 did not show any expression using qPCR or droplet digital PCR. Computational identification of drug candidates to treat keloids

[0163] Using the approach described above, compounds were identified that reverse (reduce) the expression of mRNA transcripts that are elevated during the transformation of N cells to KD cells. Additionally, compounds were identified that reverse (increase) the expression of mRNA transcripts that are reduced after the transformation of the N cells to KD cells. This generated two lists: (1 ) a list of drugs that decrease the expression of the upregulated transcripts in the KD signature across thousands of cell lines, and (2) a list of drugs that increase the expression of the downregulated transcripts, an indication that the drug might reverse the signature seen in keloid cells. The two lists were combined and the number of times that each drug was predicted to reverse the signature in the different cell lines was counted (the more times the drug tended to reverse the signature across different cell lines, the more likely it is that the drug will reverse the signature seen in keloid cells).

[0164] The compounds were ranked based on the number of instances that each drug appears in list 2 described above, and the final ranking of compounds is shown in Figure 7. Drug candidates of the 4-anilinoquinazoline family (including dasatinib, nilotinib and bosutinib) were predicted to be the most effective for treating keloids. Thus, further validation studies were performed on 4- anilinoquinazolines that have been identified as Bcr-Abl kinase inhibitors.

[0165] Cytotoxicity Assays / Half-maximal inhibitory concentration (IC50)

[0166] Previously frozen fibroblast stocks of KD (n=4) and N (n=3) were obtained for cytotoxicity evaluation of drug compounds GNF-2, nilotinib, magnolol, bosutinib and axitinib (Table 5 and Figure 8). GNF- 2 showed a significantly lower inhibitory concentration (21 ,08 ± 3,49 pM, P-value = < 0,0001 ) in KD fibroblasts than N fibroblasts, indicating that it is more potent / effective against KD than N fibroblasts. Other compounds, nilotinib, magnolol and bosutinib did not show a significant difference, and axitinib did not show any cytotoxicity in both KD and N fibroblasts.

[0167] Table 5: Cytotoxicity Assays / Half-maximal inhibitory concentration (IC50) of selected compounds to screen for efficacy in KD and N fibroblasts. Checkerboard assay for possible drug combinations

[0168] To evaluate possible drug combinations for efficacy, checkerboard assays were performed using two drugs at a time. Preliminary results for a combination of nilotinib with GNF-2 showed synergism of the two drugs with a FIC index of 0.444 (Table 6). This suggests the possibility of using drug combinations in the treatment for KD.

[0169] Table 6: Preliminary IC5o values for testing drug combinations in KD fibroblasts

[0170] Discussion

[0171] 243 upstream regulator kinases that were likely to drive the KD phenotype were identified. The list of kinases were enriched for pathway terms related to MAPK pathways, growth factors, events in mediating inflammatory events and cell cycle related events. The underlying genes for the significantly enriched KEGG and Reactome pathways were predominantly the MAPK genes and other serine / threonine protein kinase families. This is interesting because protein-serine / threonine kinases (EC 2.7.11.-) are the leading drug targets in clinical trials. Additionally, the MAPK pathway genes support cancer cell survival, which may indicate that MAPK pathway genes carry out similar mechanisms in KD pathogenesis, for their resistance to multiple treatment modalities and the high recurrence rates after surgical excision.

[0172] The Barabasi-Albert model that identified 1399 viable drug targets had 20 targets which overlapped with the X2K model. Further interrogation of the common protein kinases / biomarkers facilitated the ability to track and identify EPHA8 and MAPK12 as the key KD specific biomarkers perpetually dysregulated during the wound healing process using sequential time point analysis (4D in vivo human model).

[0173] EPHA8 belongs to the erythropoietin-producing hepatocellular receptors (Ephs). There are 14 Eph receptors in human beings which constitute class A (EPHA 1-8, EPHA10) and class B EphB1-4, EPHB6) and their ephrins (ephrin-A1 -5, ephrin-B1 -3). Ephs and their ephrin ligands are the largest subfamily of receptor tyrosine kinases (RTKs). RTKs are membrane bound proteins with the capability of initiating both forward signaling, and reverse signalling and play pivotal roles in signal transfer between neighbouring cells. It is noteworthy that amongst the Ephs and their ephrin ligands, EPHA8 fails to be expressed in other healthy human tissues. However, it exhibits low expression in healthy human skin. Dysregulated Ephs / ephrins can act as oncogenes in different cancers and they have been implicated as culprits for driving development and progression of cancer. EPHA8, reported in several cancers, was significantly upregulated on days 0 and 28 of the wound healing model (Tables 7a, b and c). Therefore, EPHA8 is a plausible driver for KD pathogenesis.

[0174] Table 7a: Key biomarkers for KD pathogenesis which inform on the drug repurposing options in addition to the enrichment analysis (Day 0)

[0175] Table 7b: Key biomarkers for KD pathogenesis which inform on the drug repurposing options in addition to the enrichment analysis (Day 14)

[0176] Table 7b: Key biomarkers for KD pathogenesis which inform on the drug repurposing options in addition to the enrichment analysis (Day 14) EPHA8 regulates cell cycle regulators, angiogenesis (Cyclin A, Cyclin D1 , CDK4) and cell migration / invasion associated genes / W / WP2 and MMP9, which are implicated in cancer metastasis. CDK4 was one of the top 5 candidates identified by utilizing the X2K model. This further supports the notion that EPHA8 is a key driver for KD pathogenesis by interacting with CDK4 during cell cycle. EPHA8 does not appear to have previously been directly implicated in skin cancers, although other Ephs like EPHB4, EPHA2, and EPHA3 have been correlated with melanoma tumour progression. This study identified several enriched Eph-ephrin signalling pathways from the X2K, Barabasi-Albert models as well as their overlapping 20 common targets. This suggests that Eph-ephrin mechanisms play a role in KD pathogenesis and further supports the Applicant’s findings that EPHA8 plays a role in KD pathogenesis.

[0177] Several ADAM (ADAM12, 23, and ADAMDEC1 ) and ADAMTS (ADAMTS20, 16, 18, 6, 4, 12, 14, 2, 1 , 9, and ADAMTSL 1) were identified. ADAMTS20 was the top significantly upregulated gene. Days 14 and 28 continued to show dysregulated ADAMTS20. The upregulated EPHA8 in this study may have some interactions with the ADAM and ADAMTS family of enzymes to drive KD pathogenesis. The observed ADAMTS expressions patterns could implicate EPHA8 as the key upstream regulator in KD pathogenesis.

[0178] The top 4 enriched Reactome pathways all had at least 3 or more underlying genes from the ADAMTS family of enzymes (ADAMTSL1, ADAMTS2, 3, 1 , 9, 10, 12, 14, 15, 16, 18, and 19) as drivers for formation of collagen or extracellular matrix organization events. Similarly, the KD-vs-N- dO Reactome pathway terms also had 2 or more of the ADAMTS family of enzymes in the top four enriched pathways. Notably, for the common KD-vs-N-dO and NS-vs-N-dO pathway terms related to ECM / collagen events, very few ADAMTS genes were observed as underlying genes for NS-vs-N- dO.

[0179] The characteristic KD invasiveness into healthy adjacent skin may be driven by upregulated EPHA8 in KD.

[0180] Upregulated EPHA8 could explain the high recurrence rates following surgical excision and the overall poor clinical outcome regardless of the different treatment modalities for KD. Glucuronidation and Pentose and glucuronate interconversions were significantly enriched in the Reactome and KEGG pathways, respectively. The UDP-glucuronosyltransferase 1 -10 enzymes were identified as key drivers for the mentioned pathways with at least 8 of UDP-Glucuronosyltransferase super family members as underlying genes for these pathway terms. These enzymes were significantly upregulated on day 14 and day 28 of the wound healing process. This family of enzymes are involved in detoxification of drugs. This could explain the poor therapeutic treatment for KD as these enzymes are involved in elimination of xenobiotics. KD treatment / management which targets these metabolic pathways on day 14 and / or day 28 after cutaneous injuries could result in better therapeutic outcomes.

[0181] The enrichment analysis from the X2K model targets identified several Eph-ephrin associated signalling events in the Reactome pathway terms. Interestingly, the underlying genes for these terms included EPHA8. Moreover, EPHA8 was identified as the underlying gene for 5 of the Eph-ephrin associated signalling events. This suggests similar mechanisms to those in tumour vasculature and angiogenesis, which supports the KD phenotype from a neoplastic perspective. Based on the sequential time points for KD-vs-N-dO, EPHA8 was significantly upregulated in KD, which correlated with upregulation of both p38 MARK isoforms MAPK12 and MAKP11. This could imply that there was a direct / indirect link between EPHA8 expression and that of p38 isoforms. The only pathway term which had both EPHA8 and MAPK12 as the underlying genes was developmental biology. EPHA8 was one of the underlying genes in axon guidance and regulation of MAPK cascade along with other MAPK genes. Interestingly, on day 14 there was no statistically significant dysregulation of EPHA8, and the same was seen for p38 isoforms. However, for KD-vs-N-d28, although EPHA8 was significantly upregulated (logFC = 1.86, AdjP = 6.73 x 103), this was a lower upregulation relative to day 0 (logFC = 3.14, AdjP = 8.62 x 104) expression patterns, which may point to the significant downregulation of MAPK11 and MAPK12.

[0182] Stress- activated protein kinase-3 / Mitogen-activated protein kinase-12 (p38y / MAPK12) is one of the four isoforms that make up the p38 MAPK family. The four isoforms are p38 / 3 / MAPK11 , p38y / MAPK12, p385 / MAPK13, and p38a / MAPK14. These isoforms share a variable sequence homology and activities. Additionally, based on their protein sequence, chemical inhibitory sensitivities and substrate specificity, they can be subdivided into p38a / p38fi and p38y / p385. The p38 MAPK family form part of the p38 pathway in the MAPK pathways. The MAPK pathways constitutes four main modules / cascades, namely: the p38 pathway, the extracellular-signal- regulated kinase 1 and 2 (ERK1 / 2), the c-Jun N-terminal and the ERK5 pathways. The initiation for transmission of extracellular signals are specific to each module. The stimuli lead to a sequential phosphorylation and activation of the respective modules following successive activation of the MAPK kinase kinase (MAPKKK), which activates the MAP kinase kinase (MAPKK) which in turn activates the MAP kinase MAPK). The activated MAPK activates diverse substrates to control many cellular processes like proliferation, differentiation, development, survival, stress response and apoptosis.

[0183] Several VEGF related events were significantly dysregulated along with multiple MAPK pathways. This may indicate cancer mechanisms are involved KD characteristic invasion into adjacent healthy skin. Interestingly the MAPK12, was often the underlying gene in these pathways. Additionally, the 20 common genes between the X2K and Barabasi-Albert models constituted several MARK pathway genes (MAPK11, MAP4K4, MAPK12).

[0184] Sequential time point analysis revealed that on days 14 and 28 of the wound healing process, glucuronidation and metabolism of xenobiotics were some of the key dysregulated terms. The combination of MARK genes and UDP-glucuronosyltransferases could be drivers for KD pathogenesis. The discovery of UDP-glucuronosyltransferases on days 14 and 28 of the wound healing process could therefore provide a timeline for therapeutic intervention.

[0185] Biopsies taken from non-lesional sites from KD prone participants have been reported to develop KD after about six weeks post biopsy. In this study, MARK signaling pathway was found dysregulated on day 0 (> 1 -year old KD scar (also referred to as a mature KD scar)) for both the KEGG and Reactome pathways. This suggests that by week 6 of the healing process in KD, the baseline day 0 (> 1 -year old KD scar) perturbations recur. KD treatment could thus be a timed procedure, with the ideal times being from about days 14-28 post wounding by targeting phase II metabolic pathways.

[0186] COMP was significantly upregulated on day 0 and day 28 supporting the narrative that cartilage-like processes are involved in KD pathogenesis.

[0187] MAPK12 was identified as a biomarker for KD pathogenesis. It is possible that the same underlying mechanisms for MAPK12 in rheumatoid arthritis also happen in KD pathogenesis for the inflammatory and cartilage / bonelike phenotype. Additionally, this study identified significantly enriched KEGG pathway term rheumatoid arthritis to further support the cartilage-like composition and the phenotype towards cartilage / bonelike phenotype.

[0188] Several KEGG pathways including PI3K-Akt signalling and mTOR signalling pathways were found to be significantly enriched / dysregulated. The list of significantly dysregulated pathways included thermogenesis and several viral infection related pathways human cytomegalovirus infection, human papillomavirus infection and human immunodeficiency virus 1 infection). Notably, all these dysregulated pathways were common for transcriptomics and proteomics enrichment analyses. Viral mechanisms in KD pathogenesis have been hypothesised before. However, no viral particles were detected in KD tissue. Upregulated Toll-like receptors (TLRs) are known to be triggered by viral infections and as a result investigation for upregulated TLRs in KD tissue were reported although the results were deemed inconclusive.

[0189] The viral prediction of KD pathogenesis that was seen in both the KEGG and Reactome pathways using the GAGE algorithm suggests that this could be an area for interrogation into KD aetiology. This is supported by the enrichment analyses from mRNA transcription data for X2K targets, the Barabasi-Albert model targets, their 20 common targets and the DEGs from KD-vs-N-dO, which resulted in several viral related events. This suggests interrogating viral mechanisms in KD from a neoplastic view. The translation is a focal point for facilitating tumour growth. Cancer cells “hijack” normal translational mechanisms to drive typical cancer behaviour. This could also suggest that from a neoplastic viewpoint KD, like cancer cells, utilizes alternative modes of translation initiation mechanisms to synthesize proteins and cell growth. This could be due to the “internal ribosome entry site” (IRES) where 40S ribosomes can bind to mRNA structures without the standard route of capdependent translation initiation. This is a known mechanism by which viral infections “hijack” normal cell processes. This could thus explain why most perturbed pathway terms from the proteomics analysis included viral mRNA translation, formation of a pool of free 40S subunits and several pathways involving mRNA translation (recruitment of ribosomes, scanning, initiation elongation, and termination). All these were significantly dysregulated Reactome pathway terms that were common across all KD group pairwise analyses. Consistent with neoplastic diseases where selenocysteine machinery is dysregulated in cancer, the findings of this study identified selenocysteine synthesis as significantly dysregulated in the KD group analyses.

[0190] The relaxin signalling pathway has been reported as one of the key altered pathways in non-small cell lung cancer. The relaxin signalling pathway was one of the top two significantly enriched KEGG pathway terms for all KD pairwise analyses (KD vs skin, KD vs scar and KD vs FKN) for proteomics analyses. This concurred with the enrichment analyses from the mRNA transcription data. Altogether the relaxin signalling pathway was enriched at both mRNA and protein levels, which suggests that it is a key pathway in KD pathogenesis. Both the relaxin and PI3K-Akt signalling pathways were dysregulated in the integrated proteomics and transcriptomics studies, suggesting that relaxin could enhance the PI3K-Akt signalling pathway in driving KD pathogenesis. Indeed, pathview visualisation of the relaxin signalling pathway depicted a link via the RXFP1 through GNAI2 / GNB2 to PI3K-Akt signalling pathway. Additionally, from the mRNA transcription data RXFP1 was significantly upregulated for KD-vs-N-dO.

[0191] Relative to the FKN group, glutathione (GSH) metabolism was significantly downregulated in the KD group. Glutathione deficiency has been reported in several diseases such as cancer, cystic fibrosis, and liver disease. In the present study, glutathione metabolism was indeed downregulated in the KD group compared to the FKN group.

[0192] The L1000 model used to identify plausible drugs for KD treatment identified several 4- anilinoquinazolines. Some of these compounds, such as dasatinib and bosutinib, were identified as Bcr-Abl kinase inhibitors and have been documented to inhibit various multiple tyrosine kinases including BCR-ABL1 , c-KIT, EPHA2, platelet-derived growth factor receptor-b and the SRC family of kinases. Interestingly, one of the key dysregulated KD genes was EPHA8 of the erythropoietinproducing hepatocellular receptors family. These findings support the use of a computational reversal gene expression approach to identify new drug candidates that can be used to treat keloids.

[0193] The results indicate that GNF-2 alone is more potent and specific to KD fibroblasts (21 ,08 ± 3,49 pM) while N fibroblasts are more tolerant to GNF-2 (48,14 ± 4,12). This indicates that GNF-2 could be used as an alternative keloid treatment to the gold standard corticosteroid, dexamethasone. Cytotoxicity assays showed that the GNF-2 IC5o (21 ,08 ± 3,49 pM) was considerably lower than the IC5o for dexamethasone (1.0 mM), suggesting that GNF-2 is more potent to KD at lower concentrations.

[0194] Conclusions

[0195] The results of this study support the use of GNF-2 or a combination of GNF-2 and a 4- anilinoquinazoline such as nilotinib as a treatment for keloids.

[0196] This study also showed several pathways including Eph-ephrin signalling and axon guidance that are associated with cancer progression and aggressive cancer grades were significantly enriched in the mRNA transcription data from this study, with EPHA8 and Eph-ephrin family members as the underlying genes. In particular, the data suggests a relationship between EPHA8 and MAPK12 for driving KD pathogenesis. The most significantly enriched pathways in the mRNA transcription data were MARK pathways. The MARK pathway genes including MAPK12 found as underlying genes in the enrichment analysis could play key roles in the unsatisfactory KD treatment outcomes and high recurrence rates. Moreover, the day 14 and day 28 time points were enriched for enriched for glucuronidation and pentose and glucuronate interconversions with UGTs superfamily members as the key drivers for these pathways. UGTs are known detoxification of drugs, suggesting their involvement in poor KD therapeutic treatment. Multiple enriched MAP kinase signaling was found on day 0 (> 1 -year old KD scar).

[0197] Therefore, a multifaceted treatment regimen and a timed process could be beneficial for treating KD, such as interventions being on about days 0, 14 and 28 (and optionally also at about 6 weeks).

[0198] A proposed treatment protocol for such a multifaceted and timed treatment is outlined in Figure 1 and described in more detail below. Non-aggressive KD scars

[0199] • On the first day of treatment (day 0), typically the day of removal of a keloid scar, a second generation 4-anilinoquinazoline such as dasatinib, bosutinib, axitinib or nilotinib is administered to the patient in combination with an allosteric inhibitor such as GNF-2, GNF-5 or ON012380.

[0200] • On about days 14 and 28, a combination of (i) a UGT inhibitor such as magnolol and (ii) dasatinib, bosutinib, axitinib or nilotinib is administered to the patient.

[0201] • At about 6 weeks, (i) a compound that targets MARK pathways, such as vemurafenib, dabrafenib, trametinib, cobimetinib or binimetinib, and / or (ii) a compound that inhibits P13K- Akt, such as idelalisib or ibrutinib is administered to the patient.

[0202] Aggressive or recurring KD scars

[0203] • On the first day of treatment (day 0), typically the day of removal of a keloid scar: o a combination of (i) a pan-Bcr-Abl inhibitor such as ponatinib and (ii) an allosteric inhibitor such as GNF-2, GNF-5 or ON012380 is administered to the patient; or o (i) a compound that inhibits MARK pathways, such as vemurafenib, dabrafenib, trametinib, cobimetinib or binimetinib, or (ii) a compound that inhibits P13K-Akt, such as idelalisib or ibrutinib, or a combination of (i) and (ii), is administered to the patient; or o a combination of (i) a mRTKI, such as lapatinib, sunitinib, sorafenib or pazopanib, and (ii) an allosteric inhibitor, such as GNF-2, GNF-5 or ON012380, is administered to the patient.

[0204] • On about days 14 and 28, a combination of (i) a UGT inhibitor such as magnolol and (ii) dasatinib, bosutinib, axitinib or nilotinib is administered to the patient.

[0205] • At about 6 weeks; o o a compound that targets MARK pathways, such as vemurafenib, dabrafenib, trametinib, cobimetinib or binimetinib, is administered to the patient; or o o a combination of (i) a mRTKI, such as lapatinib, sunitinib, sorafenib or pazopanib, and (ii) an allosteric inhibitor, such as GNF-2, GNF-5 or ON012380, 20 is administered to the patient.

[0206] The foregoing description has been presented for the purpose of illustration; it is not intended to be exhaustive or to limit the technology to the precise forms disclosed. Persons skilled in the relevant art can appreciate that many modifications and variations are possible in light of the above disclosure. The language used in the specification has been principally selected for readability and instructional purposes, and it may not have been selected to delineate or circumscribe the inventive subject matter. It is therefore intended that the scope of the present disclosure be limited not by this detailed description, but rather by any claims that issue on an application based hereon. Accordingly, the present disclosure is intended to be illustrative, but not limiting, of the scope of any accompanying claims.

[0207] Finally, throughout the specification and any accompanying claims, unless the context requires otherwise, the word ‘comprise’ or variations such as ‘comprises’ or ‘comprising’ will be understood to imply the inclusion of a stated integer or group of integers but not the exclusion of any other integer or group of integers.

[0208] References

[0209] 1 . Schuierer, S., et aL, A comprehensive assessment of RNA-seq protocols for degraded and low-quantity samples. BMC Genomics, 2017. 18(1 ): p. 442.

[0210] 2. Hedegaard, J., et aL, Next-generation sequencing of RNA and DNA isolated from paired fresh-frozen and formalin-fixed paraffin-embedded samples of human cancer and normal tissue. PLoS One, 2014. 9(5): p. e98187.

[0211] 3. Smith, P.K., et aL, Measurement of protein using bicinchoninic acid. Anal Biochem, 1985. 150(1 ): p. 76-85.

[0212] 4. Chen, E.Y., et aL, Expression2Kinases: mRNA profiling linked to multiple upstream regulatory layers. Bioinformatics, 2012. 28(1 ): p. 105-1 1.

[0213] 5. Berger, S.L, J.M. Posner, and A. Ma'ayan, Genes2Networks: connecting lists of gene symbols using mammalian protein interactions databases. BMC Bioinformatics, 2007. 8: p. 372.

[0214] 6. Lachmann, A. and A. Ma'ayan, KEA: kinase enrichment analysis. Bioinformatics, 2009. 25(5): p. 684-6.

[0215] 7. Albert, FL, H. Jeong, and A.L. Barabasi, Error and attack tolerance of complex networks. Nature, 2000. 406(6794): p. 378-82.

[0216] 8. Jeong, H., et aL, Lethality and centrality in protein networks. Nature, 2001 . 411 (6833): p. 41 - 2.

[0217] 9. Robinson, M.D. and G.K. Smyth, Moderated statistical tests for assessing differences in tag abundance. Bioinformatics, 2007. 23(21): p. 2881 -7.

[0218] 10. Koleti, A., et aL, Data Portal for the Library of Integrated Network-based Cellular Signatures (LINCS) program: integrated access to diverse large-scale cellular perturbation response data. Nucleic Acids Res, 2018. 46(D1 ): p. D558-d566. 1. Duan, Q., et aL, L1000CDS(2): LINCS L1000 characteristic direction signatures search engine. NPJ Syst Biol Appl, 2016. 2: p. 16015-. 2. Mariani, D., et aL, Antitumoral synergism between a copper(ll) complex and cisplatin improves in vitro and in vivo anticancer activity against melanoma, lung and breast cancer cells. Biochim Biophys Acta Gen Subj, 2021. 1865(10): p. 129963. 3. Aston, W.J., et aL, A systematic investigation of the maximum tolerated dose of cytotoxic chemotherapy with and without supportive care in mice. BMC Cancer, 2017. 17(1 ): p. 684. 4. Kuleshov, M.V., et aL, Enrichr: a comprehensive gene set enrichment analysis web server 2016 update. Nucleic Acids Res, 2016. 44(W1 ): p. W90-7. 5. Chen, E.Y., et aL, Enrichr: interactive and collaborative HTML5 gene list enrichment analysis tool. BMC Bioinformatics, 2013. 14: p. 128. 6. Luo, W., et aL, GAGE: generally applicable gene set enrichment for pathway analysis. BMC Bioinformatics, 2009. 10: p. 161. 7. Luo, W., et aL, Pathview Web: user friendly pathway visualization and data integration. Nucleic Acids Res, 2017. 45(W1 ): p. W501 -w508. 8. Detappe, A., Nguyen, H.VT., Jiang, Y. et ai. Molecular bottlebrush prodrugs as mono- and triplex combination therapies for multiple myeloma. Nat. Nanotechnol. 18, 184-192 (2023). htips: / / doi.org / 10.1038 / S41565-022-01310-1

Claims

CLAIMS:

1. A breakpoint cluster region-Abelson (Bcr-Abl) inhibitor for use in a method of treating a skin condition or disease in a subject.

2. A Bcr-Abl inhibitor for use according to claim 1 , wherein the skin condition or disease is keloid disease (KD) or folliculitis keloidalis nuchae (FKN).

3. A Bcr-Abl inhibitor for use according to claim 1 , wherein the skin condition or disease is keloid disease (KD).

4. A Bcr-Abl inhibitor for use according to any one of claims 1 to 3, wherein the Bcr-Abl inhibitor is GNF-2 or ON012380, or an analog, derivative or salt thereof.

5. A Bcr-Abl inhibitor for use according to claim 4, wherein the GNF-2 analog is GNF-5.

6. A Bcr-Abl inhibitor for use according to any one of claims 1 to 4, wherein the Bcr-Abl inhibitor is GNF-2.

7. A Bcr-Abl inhibitor for use according to any one of claims 1 to 6, wherein the Bcr-Abl inhibitor is administered in combination with a 4-anilinoquinazoline.

8. A Bcr-Abl inhibitor for use according to claim 7, wherein the 4-anilinoquinazoline is selected from the group consisting of dasatinib, nilotinib, bosutinib, axitinib, ponatinib, rociletinib, vemurafenib, dabrafenib, trametinib, cobimetinib, binimetinib, lapatinib, sunitinib, sorafenib and pazopanib, or a derivative or salt thereof.

9. A Bcr-Abl inhibitor for use according to claim 7, wherein the 4-anilinoquinazoline is selected from the group consisting of dasatinib, nilotinib, bosutinib and axitinib, or a derivative or salt thereof.

10. A Bcr-Abl inhibitor for use according to claim 9, wherein the 4-anilinoquinazoline is nilotinib.

11. A Bcr-Abl inhibitor for use according to claim 7, wherein the 4-anilinoquinazoline is a pan-Bcr- Abl inhibitor selected from ponatinib and rociletinib, or a derivative or salt thereof.

12. A Bcr-Abl inhibitor for use according to claim 7, wherein the 4-anilinoquinazoline is a MARKpathway inhibitor selected from the group consisting of vemurafenib, dabrafenib, trametinib, cobimetinib and binimetinib, or a derivative or salt thereof.

13. A Bcr-Abl inhibitor for use according to claim 7, wherein the 4-anilinoquinazoline is a P13K-Akt inhibitor selected from idelalisib and ibrutinib, or a derivative or salt thereof.

14. A Bcr-Abl inhibitor for use according to claim 7, wherein the 4-anilinoquinazoline is a multi receptor tyrosine kinase inhibitor (mRTKI) selected from the group consisting of lapatinib, sunitinib, sorafenib and pazopanib, or a derivative or salt thereof.

15. A Bcr-Abl inhibitor for use according to any one of claims 7 to 14, wherein the Bcr-Abl inhibitor and 4-anilinoquinazoline are administered at the same time or sequentially.

16. A Bcr-Abl inhibitor for use according to any one of claims 1 to 15, wherein the method of treatment further comprises administering a UDP-glucuronosyltransferase (UGT) inhibitor such as magnolol or a derivative thereof to the subject.

17. A Bcr-Abl inhibitor for use according to claim 16, wherein the UGT inhibitor is administered about one to four weeks after administration of the Bcr-Abl inhibitor.

18. A Bcr-Abl inhibitor for use according to any one of claims 1 to 17, wherein the method of treatment further comprises administering a 4-anilinoquinazoline to the subject about one to four weeks after administration of the Bcr-Abl inhibitor.

19. A Bcr-Abl inhibitor for use according to any one of claims 16 to 18, wherein the UGT and / or 4- anilinoquinazoline are administered about 2 weeks after administration of the Bcr-Abl inhibitor.

20. A Bcr-Abl inhibitor for use according to any one of claims 16 to 19, wherein the UGT and / or 4- anilinoquinazoline are administered about 4 weeks after administration of the Bcr-Abl inhibitor.21 . A Bcr-Abl inhibitor for use according to any one of claims 1 to 20, wherein the method of treatment further comprises administering one or more 4-anilinoquinazolines to the subject about six weeks after administration of the Bcr-Abl inhibitor.

22. A Bcr-Abl inhibitor for use according to claim 21 , wherein a 4-anilinoquinazoline that is administered about six weeks after administration of the initial Bcr-Abl inhibitor is selected from the group consisting of vemurafenib, dabrafenib, trametinib, cobimetinib and binimetinib, or aderivative or salt thereof.

23. A Bcr-Abl inhibitor for use according to claim 21 or 22, wherein a 4-anilinoquinazoline that is administered about six weeks after administration of the initial Bcr-Abl inhibitor is selected from the group consisting of idelalisib or ibrutinib, or a derivative or salt thereof.

24. A Bcr-Abl inhibitor for use according to claim 21 , wherein a 4-anilinoquinazoline that is administered about six weeks after administration of the initial Bcr-Abl inhibitor is selected from the group consisting of lapatinib, sunitinib, sorafenib and pazopanib, or a derivative or salt thereof.

25. A Bcr-Abl inhibitor for use according to claim 24, wherein a Bcr-Abl inhibitor selected from the group consisting of GNF-2, GNF-5 and ON012380 is administered with the 4- anilinoquinazoline that is administered about six weeks after administration of the initial Bcr- Abl inhibitor.

26. A combination of a Bcr-Abl inhibitor and a 4-anilinoquinazoline for use in a method of treating a skin condition or disease, wherein: the Bcr-Abl inhibitor is selected from the group consisting of GNF-2, GNF-5 and ON012380, or a derivative or salt thereof; and the 4-anilinoquinazoline is selected from the group consisting of dasatinib, nilotinib, bosutinib, axitinib, ponatinib, rociletinib, vemurafenib, dabrafenib, trametinib, cobimetinib, binimetinib, lapatinib, sunitinib, sorafenib and pazopanib, or a derivative or salt thereof.

27. A combination for use of claim 26, wherein the Bcr-Abl inhibitor is GNF-2 and the 4- anilinoquinazoline is nilotinib.

28. Use of a breakpoint cluster region-Abelson (Bcr-Abl) inhibitor in the manufacture of a medicament for use in treating a skin condition or disease in a subject.

29. The use according to claim 28, wherein the skin condition or disease is keloid disease (KD) or folliculitis keloidalis nuchae (FKN).

30. The use according to claim 28, wherein the skin condition or disease is keloid disease (KD).31 . The use according to any one of claims 28 to 30, wherein the Bcr-Abl inhibitor is GNF-2 or ON012380, or an analog or derivative thereof.

32. The use according to claim 31 , wherein the GNF-2 analog is GNF-5.

33. The use according to any one of claims 28 to 31 , wherein the Bcr-Abl inhibitor is GNF-2.

34. The use according to any one of claims 28 to 33, wherein the Bcr-Abl inhibitor is administered with a 4-anilinoquinazoline.

35. The use according to claim 34, wherein the 4-anilinoquinazoline is selected from the group consisting of dasatinib, nilotinib, bosutinib, axitinib, ponatinib, rociletinib, vemurafenib, dabrafenib, trametinib, cobimetinib, binimetinib, lapatinib, sunitinib, sorafenib and pazopanib, or a derivative or salt thereof.

36. The use according to claim 34, wherein the 4-anilinoquinazoline is selected from the group consisting of dasatinib, nilotinib, bosutinib and axitinib, or a derivative or salt thereof.

37. The use according to claim 36, wherein the 4-anilinoquinazoline is nilotinib or a salt thereof.

38. The use according to claim 34, wherein the 4-anilinoquinazoline is a pan-Bcr-Abl inhibitor selected from ponatinib and rociletinib, or a derivative or salt thereof.

39. The use according to claim 34, wherein the 4-anilinoquinazoline is a MARK pathway inhibitor selected from the group consisting of vemurafenib, dabrafenib, trametinib, cobimetinib and binimetinib, or a derivative or salt thereof.

40. The use according to claim 34, wherein the 4-anilinoquinazoline is a P13K-Akt inhibitor selected from idelalisib and ibrutinib, or a derivative or salt thereof.41 . The use according to claim 34, wherein the 4-anilinoquinazoline is a multi receptor tyrosine kinase inhibitor (mRTKI) selected from the group consisting of lapatinib, sunitinib, sorafenib and pazopanib, or a derivative or salt thereof.

42. The use according to any one of claims 34 to 41 , wherein the Bcr-Abl inhibitor and 4- anilinoquinazoline are administered at the same time or sequentially.

43. The use according to any one of claims 28 to 42, wherein the method of treatment further comprises administering a UDP-glucuronosyltransferase (UGT) inhibitor such as magnolol ora derivative or salt thereof to the subject.

44. The use according to claim 43, wherein the UGT inhibitor is administered about one to four weeks after administration of the Bcr-Abl inhibitor.

45. The use according to any one of claims 28 to 44, wherein the method of treatment further comprises administering a 4-anilinoquinazoline to the subject about one to four weeks after administration of the Bcr-Abl inhibitor.

46. The use according to any one of claims 43 to 45, wherein the UGT and / or 4-anilinoquinazoline are administered about 2 weeks after administration of the Bcr-Abl inhibitor.

47. The use according to any one of claims 43 to 46, wherein the UGT and / or 4-anilinoquinazoline are administered about 4 weeks after administration of the Bcr-Abl inhibitor.

48. The use according to any one of claims 28 to 47, wherein the method of treatment further comprises administering one or more 4-anilinoquinazolines to the subject about six weeks after administration of the initial Bcr-Abl inhibitor.

49. The use according to claim 48, wherein a 4-anilinoquinazoline that is administered about six weeks after administration of the initial Bcr-Abl inhibitor is selected from the group consisting of vemurafenib, dabrafenib, trametinib, cobimetinib and binimetinib, or a derivative or salt thereof.

50. The use according to claim 48 or 49, wherein a 4-anilinoquinazoline that is administered about six weeks after administration of the initial Bcr-Abl inhibitor is selected from the group consisting of idelalisib or ibrutinib, or a derivative or salt thereof.51 . The use according to claim 48, wherein a 4-anilinoquinazoline that is administered about six weeks after administration of the initial Bcr-Abl inhibitor is selected from the group consisting of lapatinib, sunitinib, sorafenib and pazopanib, or a derivative or salt thereof.

52. The use according to claim 51 , wherein a Bcr-Abl inhibitor selected from the group consisting of GNF-2, GNF-5 and ON012380, or a derivative or salt thereof, is administered with the 4- anilinoquinazoline that is administered about six weeks after administration of the initial Bcr- Abl inhibitor.

53. A method of treating a skin condition or disease in a subject, the method comprising administering an effective amount of a breakpoint cluster region-Abelson (Bcr-Abl) to the subject.

54. The method of claim 53, wherein the skin condition or disease is keloid disease (KD) or folliculitis keloidalis nuchae (FKN).

55. The method of claim 53, wherein the skin condition or disease is keloid disease (KD).

56. The method of any one of claims 53 to 55, wherein the Bcr-Abl inhibitor is GNF-2 or ON012380, or an analog or derivative thereof.

57. The method of claim 56, wherein the GNF-2 analog is GNF-5.

58. The method of any one of claims 53 to 56, wherein the Bcr-Abl inhibitor is GNF-2.

59. The method of any one of claims 53 to 58, wherein the Bcr-Abl inhibitor is administered with a4-anilinoquinazoline.

60. The method of claim 59, wherein the 4-anilinoquinazoline is selected from the group consisting of dasatinib, nilotinib, bosutinib, axitinib, ponatinib, rociletinib, vemurafenib, dabrafenib, trametinib, cobimetinib, binimetinib, lapatinib, sunitinib, sorafenib and pazopanib, or a derivative or salt thereof.61 . The method of claim 59, wherein the 4-anilinoquinazoline is selected from the group consisting of dasatinib, nilotinib, bosutinib and axitinib, or a derivative or salt thereof.

62. The method of claim 61 , wherein the 4-anilinoquinazoline is nilotinib, or a salt thereof.

63. The method of claim 59, wherein the 4-anilinoquinazoline is a pan-Bcr-Abl inhibitor selected from ponatinib and rociletinib, or a derivative or salt thereof.

64. The method of claim 59, wherein the 4-anilinoquinazoline is a MARK pathway inhibitor selected from the group consisting of vemurafenib, dabrafenib, trametinib, cobimetinib and binimetinib, or a derivative or salt thereof.

65. The method of claim 59, wherein the 4-anilinoquinazoline is a P13K-Akt inhibitor selected fromidelalisib and ibrutinib, or a derivative or salt thereof.

66. The method of claim 59, wherein the 4-anilinoquinazoline is a multi receptor tyrosine kinase inhibitor (mRTKI) selected from the group consisting of lapatinib, sunitinib, sorafenib and pazopanib, or a derivative or salt thereof.

67. The method of any one of claims 59 to 66, wherein the Bcr-Abl inhibitor and 4- anilinoquinazoline are administered at the same time or sequentially.

68. The method of any one of claims 53 to 67, wherein the method of treatment further comprises administering a UDP-glucuronosyltransferase (UGT) inhibitor such as magnolol or a derivative thereof or salt thereof to the subject.

69. The method of claim 68, wherein the UGT inhibitor is administered about one to four weeks after administration of the Bcr-Abl inhibitor.

70. The method of any one of claims 53 to 69, wherein the method of treatment further comprises administering a 4-anilinoquinazoline to the subject about one to four weeks after administration of the initial Bcr-Abl inhibitor.71 . The method of any one of claims 68 to 70, wherein the UGT and / or 4-anilinoquinazoline are administered about 2 weeks after administration of the initial Bcr-Abl inhibitor.

72. The method of any one of claims 68 to 71 , wherein the UGT and / or 4-anilinoquinazoline are administered about 4 weeks after administration of the initial Bcr-Abl inhibitor.

73. The method of any one of claims 53 to 72, wherein the method of treatment further comprises administering one or more 4-anilinoquinazolines to the subject about six weeks after administration of the initial Bcr-Abl inhibitor.

74. The method of claim 73, wherein a 4-anilinoquinazoline that is administered about six weeks after administration of the initial Bcr-Abl inhibitor is selected from the group consisting of vemurafenib, dabrafenib, trametinib, cobimetinib and binimetinib, or a derivative or salt thereof.

75. The method of claim 73 or 74, wherein a 4-anilinoquinazoline that is administered about six weeks after administration of the initial Bcr-Abl inhibitor is selected from the group consisting of idelalisib and ibrutinib, or a derivative or salt thereof.

76. The method of claim 73, wherein a 4-anilinoquinazoline that is administered about six weeks after administration of the initial Bcr-Abl inhibitor is selected from the group consisting of lapatinib, sunitinib, sorafenib and pazopanib, or a derivative or salt thereof.

77. The method of claim 76, wherein a Bcr-Abl inhibitor selected from the group consisting of GNF- 2, GNF-5 and ON012380 is administered with the 4-anilinoquinazoline that is administered about six weeks after administration of the initial Bcr-Abl inhibitor.

78. A kit comprising: a Bcr-Abl inhibitor selected from the group consisting of GNF-2, GNF-5 and ON012380, or a derivative or salt thereof; and a 4-anilinoquinazoline selected from the group consisting of dasatinib, nilotinib, bosutinib, axitinib, ponatinib, rociletinib, vemurafenib, dabrafenib, trametinib, cobimetinib, binimetinib, lapatinib, sunitinib, sorafenib and pazopanib, or a derivative or salt thereof.

79. The kit of claim 78, wherein the Bcr-Abl inhibitor is GNF-2 or a salt thereof and the 4- anilinoquinazoline is nilotinib or a salt thereof.

Images