Live Biotherapeutics

In recent years, probiotics have expanded from their traditional classification as “health promoting food” to the development of live biotherapeutic products (LBP). Traditional probiotics are marketed as food/dietary supplements while LBPs are drug products intended for treatment or prevention of diseases. This type of products offers several advantages over traditional drugs, but also entail potential challenges with development, manufacturing, and demonstration of clinical safety. To obtain a sufficient quality, LBPs are typically produced by cultivation in a bioreactor, followed by formulation and lyophilization. In the first part of the project, the impact of lyophilization parameters on physicochemical and biological properties of Limosilactobacillus reuteri R2LC was evaluated. Using sucrose as a lyoprotectant gave a better freeze-drying survival, vitality and storage stability than using trehalose. A high concentration (20%) of sucrose sometimes resulted in a collapsed structure and 15% gave the overall best properties of the lyophilized bacteria. Interestingly, vitality was positively affected by using a higher concentration (10 cfu/ml) of bacteria. Another observation was that introducing an annealing step in the process was positive when using sucrose as lyoprotectant, but no effect was seen when using trehalose. The second part of the project describes evaluation of the genetically modified L. reuteri R2LC expressing the human chemokine CXCL12 (ILP100-Topical) in a phase 1 trial on wound healing. The product was safe and well-tolerated. In addition, it gave a larger proportion of healed wounds (76 %) on Day 32 when compared to saline/placebo (59 %) (p=0.020) and the time of wound healing was reduced by 6 days on average and by 10 days at highest dose. Also, ILP100-Topical increased the density of CXCL12 cells in the wounds and local wound blood perfusion.

Nyckelord: Probiotika, Levande bioterapeutika, Limosilactobacillus reuteri R2LC, ILP100-Topical, frystorkning, vitalitet, viabilitet, säkerhet, sårläkning. The contribution of Nisha Tyagi to the papers I and II included in this thesis was as follows: I. Planning and designing of experimental work together with supervisors, performed most of experimental work, data analysis and manuscript writing.
II. Methodology designed together with coauthors. Method development, quality control and validation of Laser Speckle Contract Analysis Image analysis of perfusion on skin and in wounds and analysis of microcirculation of the treated wounds and surrounding skin.   The importance of the microbiota in health and disease have been described in many studies (Hou et al., 2022;Tungland, 2018). Human microbiotas are classified based on the sites they colonize, such as oral cavity, skin, gut etc. (Hou et al., 2022). The bacteria assists in metabolic processes, degradation of food components, production of vitamins, immune development, as well as protection against pathogens (Hou et al., 2022). Probiotics are defined as live microorganisms that when administered in adequate amount confer a health benefit on the host (Hill et al., 2014). Traditionally, strains used as probiotics have been isolated from the gastrointestinal tract (incl. faeces), breast milk or fermented food products (for example yogurt and kefir) and have been marketed as food, food/ dietary supplements (Fenster et al., 2019). Recent advancement in technologies such as DNA sequencing, bioinformatics and metabolomics have been strong drivers for the expansion of knowledge of the human microbiota at the molecular and strain levels. Projects such as Human Microbiome Project (https://hmpdacc.org/) and MetaHIT (https://sanger.ac.uk/resources/downloads/bacteria/metahit/) have played important roles in connecting knowledge about the gut microbiota to the broader medical research. The increased research within the field has also been a vital help in the discovery of new bacterial strains that can be used for developing potential new probiotics/ Live biotherapeutics. First generation (also called traditional) probiotics are primarily used as food or dietary supplements, while next-generation probiotics or live biotherapeutic products (LBP) are intended to use as medicinal drug products to prevent or cure human diseases (O'Toole et al., 2017). LBPs also include 1. Introduction genetically modified microorganisms (GMM) (O'Toole et al., 2017). Live Biotherapeutic Products are defined as 'a biological product that 1) contains live microorganisms, such as bacteria; 2) is applicable to the prevention, treatment, or cure of a disease or conditions of human beings; and 3) is not a vaccine' (U.S. Department of Health and Human Services, 2016). .

LBPs in treatment and prevention of human diseases
Probiotic products have been developed and distributed over the past century, while the first live biotherapeutic drugs just recently have been approved by US-FDA: Rebyota (Nov. 2022) and VOWEST (April 2023). Both products are based on a complete faecal microbiota, and thus have big similarities to the concept "faecal microbiota transplantation" (FMT; (Ooijevaar et al., 2019). Development of next-generation probiotic is an exciting and promising advance of the healthcare sector. Another FMT product from BiomeBank was approved only in Australia to treat recurrent Clostridioides difficile infection. Whilst there is plethora ongoing research only a few LBP drug candidates have reached to clinical stage being evaluated in randomized controlled trials following the pharmaceutical regulatory frameworks (described in table 2). The market of live biotherapeutic products grow quickly and globally it is anticipated to reach $2.60 billion by 2030 (Analytic, 2022).
LBPs could potentially be helpful in treatment and prevention of a wide range of diseases such as metabolic disorders, cardiovascular diseases, infectious diseases, cancer, and inflammation (Lim & Song, 2019; Meng et al., 2023). Most LBPs under development are based on naturally occurring bacterial strains (60%) but leveraging the area of synthetic biology and metagenomics are now used for the development of genetically engineered LBP's (40%) ( Table 2). Examples are Aurealis therapeutics that have developed AUP-16, a genetically engineering Lactococcus lactis expressing human fibroblast growth factor 2, IL-4 and CSF-1 to treat diabetic foot ulcers (Kurkipuro et al., 2022), and Ilya Pharma that have developed a strain of Limosilactobacillus reuteri that expresses the human chemokine CXCL12 for accelerated wound healing (paper II).

Lactic acid bacteria (LAB)
The mankind's use of lactic acid bacteria (LAB) goes long back to ancient times. LAB have played important roles in preservation of food, maintaining texture and incorporation of flavours in the food and as additive to improve gut health (Piccioni et al., 2021;Stiles, 1996). The importance of lactic acid bacteria for maintaining the gut flora and healing digestive problems was first suggested in 1907 (Metchnikoff, 1907), but the interest in the link to health has increased significantly during the last decades. The first step in the production of lactic acid bacteria/microorganisms to be used in LBPs is cultivation followed by formulation and drying. During growth, lactic acid bacteria ferment carbohydrates into end products such as lactic and acetic acids or ethanol (Stanbury et al., 2017). Based on the pathway used for sugar fermentation LAB are divided into two groups: homofermentative (lactic acid as the main end product) and heterofermentative (lactic acid, carbon dioxide, and ethanol or acetic acid as main end products). L. reuteri R2LC belongs to the latter group (Yu et al., 2018). It has previously been demonstrated that the biological activities of probiotics are directly influenced by cultivation parameters such as time of harvest, growth media, and environmental parameters (e.g. temperature and pH) (Meng et al., 2008). Freeze-drying (also known as lyophilization) is a widely used method for drying probiotics (including LBP's) to increase the shelf life of the product. The process is divided into three steps: a) freezing, b) primary drying, and c) secondary drying (Fonseca et al., 2015) ( Figure 1). During the freezing step, samples are cooled by decreasing shelf temperature resulting in ice crystal formation. Usually, the temperature is between -40°C and -20°C. During this step bacteria is exposed to osmotic shock, mechanical and oxidative stress, which can later affect the biological activity of the probiotics/LBPs. To partly overcome these stressful conditions during freezing, fast cooling and including an annealing step could be utilized, which facilitates water vapour transport and improves the drying (Merivaara et al., 2021). Annealing is a process to optimize the freezing and facilitate the primary drying i.e., lower than glass transition temperature (Tg) for certain period in order to increase the ice crystal growth and promote the rate of primary drying (Badal Tejedor et al., 2020). Previously, it has been shown that annealing improved the viability and stability of probiotics (Ekdawi-Sever et al., 2003). During the primary drying, the pressure is decreased which leads to ice sublimation, and it's important to have an optimal combination of shelf temperature and chamber pressure to achieve an efficient sublimation (Fonseca et al., 2015). In the last step, the secondary drying, unfrozen water is removed via desorption. In this step the temperature is slowly increased to around 20°C while maintaining a low pressure. , and e.g. sucrose and trehalose are well known lyoprotectants in the production of probiotics and biological drugs. It has been shown that non-reducing disaccharides protect biological drugs/probiotics by forming a glassy matrix structure that prevent damage due to ice crystal formation. These sugars also have a high glass transition temperature that stabilizes the product and prolong shelf life at higher temperature (Bodzen et al. Development of live biotherapeutic products (LBP's) entails challenges like designing an efficient product concept, development of a production methods that ensures a high and consistent quality, and demonstration of both clinical safety and efficacy. This thesis is a continuation on previous publications that describe intriguing interactions and effects of Limosilactobacillus reuteri R2LC in preclinical models (Ahl et al., 2016; Liu et al., 2021); and efficient wound healing by using R2LC that has been genetically engineered to express the chemokine CXCL12 (ILP100-Topical). The thesis consists of the following two parts: • The effect of formulation and lyophilization parameters on biological as well as physicochemical properties of freeze-dried L. reuteri R2LC. The following parameters were investigated: type of lyoprotectant, concentration of lyoprotectant, bacterial concentration and using a freeze-drying process with or without an annealing step. The following biological characteristics were monitored: process survival, vitality, and shelf life. In addition, the correlation between physicochemical properties of the lyophilized products and the biological characteristics were investigated. (Paper I).
• In a first-in-human clinical phase-I trial on wound healing, the safety, tolerability, and biologic effect of genetically modified L. reuteri R2LC (ILP100-Topical) was evaluated after topical single and multiple dose administration to experimentally induced skin wounds in healthy subjects/volunteers. Furthermore, influence of ILP100-Topical on the microcirculation/blood perfusion in the

Aims
In paper I, Limosilactobacillus reuteri R2LC was used as a model microorganism for studying the impact of freeze-drying and formulation on its performance while in paper II a genetically modified Limosilactobacillus reuteri R2LC expressing the chemokine CXCL12 (ILP100-Topical; (Vagesjo et al., 2018) was evaluated in a first-in-human phase-1 trial.
3.1 Impact of production parameters on physicochemical and biological properties of freeze-dried L. reuteri R2LC (Paper I)

Experimental design
To understand the impact of formulation and freeze-drying on physicochemical and biological properties of freeze-dried L. reuteri R2LC, four experimental factors were combined to in total 24 different variants using a Design-of-Experiment approach and a full factorial study design (using the software Modde (Eriksson et al., 2008)). Type of lyoprotectant (sucrose or trehalose); Concentration of lyoprotectant (10, 15 or 20%); Bacterial concentration (10 9 or 10 10 cfu/ml); and introducing an annealing step or not in the freeze-drying process. (Figure 2).  L. reuteri R2LC was cultivated in a pilot 5-L scale bioreactor (also called fermenter) containing deMan, Rogosa, Sharpe broth (MRS, Merck) as shown in figure 3. The stirring speed, pH, and temperature of bioreactor were set to 200 rpm, 5.7 and 37°C respectively. During fermentation the growth was monitored by measuring optical density as well as by plating the sample on MRS agar plates via serial dilution at different timepoints (1, 2, 3 hr and so on). The detailed materials and methods description is provided in paper I. After fermentation the cell suspension was concentrated using a diafiltration column (750 kDa cut off) and then mixed with different lyoprotectants at different concentrations as shown in figure 4. The formulated samples were firstly divided into sets (1) samples freeze-dried with annealing and (2) samples freeze-dried without annealing step. Two Christ, Epsilon 2-6D, LSC plus, (Martin Christ GmbH, Germany) freeze dryers were used. Detailed description of freeze-drying process can be found in paper I.

Biological Characterization
To ensure the functionality of freeze-dried probiotics/LBP's, certain criteria need to be fulfilled. It is important that the bacteria (i) are alive and have the expected concentration, (ii) have a high metabolic activity, and (iii) have a sufficient long-term stability. Therefore, the effects of the different production parameters on biological activities of freeze-dried R2LC was analysed using the different methods described below.

Bacterial viability
Monitoring of the bacterial survival after freeze-drying was done by plate count analysis of samples taken before and after the drying. The samples were plated on MRS agar plates after serial dilution (as per different bacterial concentration).

Cell vitality
The vitality of the bacterial cells was measured by a pH drop method. During cultivation R2LC produces lactic acid which reduce the pH of the media. Metabolically active bacterial cells (having high vitality) therefore give a larger drop in pH than cells with low vitality. After reconstitution of freezedried bacteria in growth media, pH drop was measured at 3 timepoints 0, 1, and 2 hours and delta pH were calculated.

Accelerated storage stability
The stability of the freeze-dried product is related to the storage temperature (Meng et al., 2008). A probiotics/LBP is normally stored at 4-25°C, but an accelerated stability study was performed at 37°C. We investigated the viability, vitality, and water content of freeze-dried L. reuteri R2LC before and after storage at 37°C for 2 and 4 weeks.

Physiochemical characterisation
The freeze-dried samples were visually analysed, and the appearance of the cakes was scored according to the following scale: Score 1) intact and homogenous cake; 2) intact but non-homogenous cake, colour change of the bottom part of the cake; 3) shrinkage of the cake around the edges; 4) partially collapsed cake (20-40%), and 5) collapsed cake. ( Figure 5). Glass transition temperature (Tg) of all L. reuteri formulations were measured by differential scanning calorimetry (DSC) as described in paper I. Apart from DSC, we also studied the distribution of freeze-dried bacteria in the lyophilized cake and porosity of different formulations, after investigating the material with scanning electron microscopy (SEM; the same samples that were analysed by DSC). Sections from top and bottom were visualized at 100 x, 500 x, 1000 x and 2500 x magnifications. Porosity of the cake was measured using the Image J software according to the procedure presented by Saraf et al. Bacterial aggregation in the lyophilized samples was measured using flow cytometry, where clumps larger than >6 µm were defined as aggregates. Prior to analysis samples were diluted 1:100 and 1:200 with saline solution. 100,000 events were recorded for all formulations. FlowJo software was used for calculating the aggregation (%). The aggregation (%) was calculated as (no. of events counted in bead region х no. of events in bacterial region)/100.
Water content of all formulations was analysed by using a Karl Fischer coulometric method. All samples were reconstituted in dry methanol and incubated at room temperature for 1 hour to extract all water. Supernatants was analysed by Metrohm 831 Karl Fischer Coulometry to determine water content.

Evaluation of ILP100 in a first-in-human phase-1 trial on wound healing (Paper II)
Paper II presents an adaptive, randomized, double-blind, placebo-controlled first-in-human study designed to evaluate safety, tolerability, clinical and biologic effects on wound healing of single and multiple ascending doses of ILP100-Topical (L. reuteri expressing CXCL12 administered topically to experimentally induced skin wounds in healthy subjects). The study comprises of a treatment and assessment phase up to 6 weeks after last treatment and a 5-year long-term follow-up.

Microcirculation/blood perfusion in wound
Microcirculation/blood perfusion around wound edges was measured using PeriCam PSI NR system from perimed, which is based on laser speckle contrast analysis (LASCA) imaging techniques that assess blood perfusion in real time. Region of interest (ROI) was defined as areas within a perfusion image where the analysis was done. The following three different ROIs were made a) the wound area, b) the area 5 mm, and c) 10 mm outside the wound area ( Figure 7). The PIMSoft software was used to assess the blood perfusion images of the wound and the surrounding. The detailed procedure is described in paper II (Öhnstedt et al., 2023).   Figure 9). In addition, we also observed that aggregation was higher when using high (10 10 CFU/mL) bacterial concentration and a positive correlation between aggregation and vitality was also seen. ( Figure 10; Table 3).

Paper II
ILP100-Topical is a genetically modified L. reuteri R2LC expressing human chemokine CXCL12-1a (designated ILP100-Topical) that has been designed to accelerate wound healing (Vagesjo et al., 2018). The chemokine CXCL12 binds CXCR4 expressed by immune cells and keratinocytes. Macrophages and neutrophils are major immune cells at the wound site, where they are important for controlling invading microorganisms and for facilitating the healing process by secreting additional chemokines, growth factors, and matrix digesting enzymes. During this process, macrophages shift phenotype and become anti-inflammatory, and subsequently promote closing of the wound. This is induced by macrophage phagocytosis of cell debris and by microenvironmental signals such as CXCL12. It has also been shown that on-site delivery of ILP100-Topical reduces the pH of the wound environment which inactivates the protease CD26 and helps in increasing the bioavailability of CXCL12. The chemokine expressing L. reuteri R2LC have previously been shown to accelerate wound healing in healthy mice, ischemic, and hyperglycaemic murine models (Vagesjo et al., 2018). The aim of this study was to determine the safety and tolerability as well as clinical and biologic effects of ILP100-Topical after topical single and multiple dose administration to experimentally induced skin wounds in healthy subjects. Results showed neither adverse events (AEs) nor colonization of the genetically modified L. reuteri R2LC in blood and faeces samples of the patients. Also, no increase of concentration of circulating CXCL12 was observed in individuals treated with single or multi-dose of ILP100-Topical. Furthermore, there was no increase in wound rupture in connection to treatment with ILP100-Topical, while two cases of wound scars were reported in the placebo group. All dataset, single as well as multidosage were considered safe and well-tolerated over 3 weeks of administered ILP100. The microcirculation was measured in the wounds and the skin surrounding the wound areas. A wound was defined as healed when the wound area was completely re-epithelialized. In multi-dosing ILP100, significant difference in treatment-related wound healing in cohort 1 at day 32 and day 19 and 22 (Day 32, p=0.058) was observed. Pooled analyses of all cohorts showed that compared to control ILP100 significantly improved wound healing (p=0.020) as shown in Figure 6 and Table 4.
Recently, the focus to accelerate wound healing by changing the wound environment by topical application of growth factors or cell therapies have been increased Mahdipour & Sahebkar, 2020). In the previous preclinical study by Vågesjö et al. (Vagesjo et al., 2018) it was demonstrated that the blood flow in hyperglycaemic mice was normalized in wounds treated with ILP100. In another study, genetically modified L. lactis expressing FGF2, CSF1, and IL-4 resulted in accelerated wound healing in a mouse model (Kurkipuro et al., 2022).  Paper I: We have observed that the factors: type of lyoprotectant, lyoprotectant concentration, bacterial concentration and annealing affect the properties and performance of freeze-dried R2LC. Sucrose as a lyoprotectant gave better freeze-drying survival, vitality, and storage stability of R2LC than trehalose as a lyoprotectant. Overall, sucrose at 15% with an annealing step showed the best results in the analyses of freeze-drying survival, vitality, and storage stability of R2LC. The high concentration of sucrose (20%) at low bacterial concentration (10 9 CFU/mL) resulted in elevated water content and resulted in partial and collapsed cake formation. Finally, the high (10 10 CFU/mL) R2LC concentration resulted in the best vitality but also resulted in more aggregates. Paper II: L. reuteri R2LC expressing CXCL12 (ILP100-Topical) has previously shown promising effects in accelerating wound healing by onsite delivery of CXCL12, enhancing the activity of macrophages. In a human clinical phase-1 trial (paper II) we showed that ILP100-Topical is safe, well tolerable, and have an effective biologic effect in accelerating wound healing. Time to first healing was shortened by 6 days on average, and by 10 days in highest dose.

Conclusions
• Study I -we have investigated the impact of different process parameters on the properties of freeze-dried L. reuteri R2LC. A future aim could be to investigate the biological properties of freezedried LBP in different model such as an animal model or an in vitro artificial small intestine (SHIME). This system provides the possibility to perform a realistic assessment of probiotic/LBP properties in an environment with extensive similarities to the gastrointestinal tract. SHIME could be operated to simulate different intestinal environments such as adult, infants, elderly, and specific conditions (e.g., pathogen infection). • Study II -The first-in-human study have demonstrated the product to be safe and effective in accelerating wound healing. There are one ongoing phase 2a trial investigating the ILP100-Topical in diabetic patients with diabetes and non-healing wounds and an IND cleared for a pivotal trial evaluating the ILP100-Topical in post-surgical wounds in prediabetic, diabetic and obese patients.

Summary
Background Impaired wound healing is a growing medical problem and very few approved drugs with documented clinical efficacy are available. CXCL12-expressing lactic acid bacteria, Limosilactobacillus reuteri (ILP100-Topical), has been demonstrated to accelerate wound healing in controlled preclinical models. In this first-in-human study, the primary objective was to determine safety and tolerability of the drug candidate ILP100-Topical, while secondary objectives included assessments of clinical and biologic effects on wound healing by traditionally accepted methods and explorative and traceable assessments.
Methods SITU-SAFE is an adaptive, randomised, double-blind, placebo-controlled, first-in-human phase 1 trial (EudraCT 2019-000680-24) consisting of a single (SAD) and a multiple ascending dose (MAD) part of three dose cohorts each. The study was performed at the Phase 1 Unit, Uppsala University Hospital, Uppsala, Sweden. Data in this article were collected between Sep 20th, 2019 and Oct 20th 2021. In total 240 wounds were induced on the upper arms in 36 healthy volunteers. SAD: 12 participants, 4 wounds (2/arm), MAD: 24 participants, 8 wounds (4/arm). Wounds in each participant were randomised to treatment with placebo/saline or ILP100-Topical.
Findings In all individuals and doses, ILP100-Topical was safe and well-tolerated with no systemic exposure. A combined cohort analysis showed a significantly larger proportion of healed wounds (p = 0.020) on Day 32 by multi-dosing of ILP100-Topical when compared to saline/placebo (76% (73/96) and 59% (57/96) healed wounds, respectively). In addition, time to first registered healing was shortened by 6 days on average, and by 10 days at highest dose. ILP100-Topical increased the density of CXCL12 + cells in the wounds and local wound blood perfusion.

Introduction
Complicated or non-healing wounds, encompassing wounds that do not heal for 4 or more weeks with standard of care, are a growing medical problem associated with metabolic diseases and aging. 1-4 These problematic wounds negatively impact life quality and reduce life expectancy, and they often become infected and increase the risk for sepsis. There is a high unmet need for effective therapies, as there are very few available therapeutics with proven efficacy of accelerated wound healing. Instead, antibiotics are being overused in these patients, and up to 75% receive systemic antibiotics despite often lacking documented clinical infection. 5,6 Wound healing is driven by cells of the immune system regulated by signals from the wound microenvironment. [7][8][9] Immunomodulatory drugs are currently transforming oncology and autoimmune diseases, while therapeutic targeting of immune cells within wounds has not yet been successful. This is at least in part due to that topical administration is hampered by the proteolytic wound environment, which limits the bioavailability of candidate therapeutic molecules. 10 Therefore, genetically modified bacteria producing, delivering, and stabilising immunomodulatory proteins within the wounds could be disruptive in the field of immunotherapy, as they enable the use of proteins with short half-lives as scalable therapeutics.
A first-in-class drug candidate, ILP100-Topical (emilimogene sigulactibac), was designed by engineering Limosilactobacillus reuteri R2LC (L. reuteri R2LC), a strain of non-human origin, to produce and release the human chemokine CXCL12-α on-site to the wound bed. 11 Accelerated healing after topical delivery has been well-documented in multiple non-clinical studies, depends on increased numbers of wound macrophages of a restorative phenotype expressing TGF-β, and a favourable safety profile was demonstrated. 5,12 Here, we present results from the randomised, blinded, and placebo-controlled first-in-human study designed to primarily assess safety and tolerability of ILP100-Topical, whereas the secondary and exploratory objectives aimed to evaluate clinical and biologic effects on wound healing. To complement and validate the conventional assessments performed by Investigators during visits, blinded and highresolution wound imaging techniques were used, which provided objective analyses of healing in fully traceable and reproducible data sets. This pioneering study demonstrates a favourable safety profile together with proven clinical and biologic effects on accelerated wound healing, which supports the continued clinical development of ILP100-Topical, a new modality and local immunotherapeutic.

Study design
This single-centre adaptive, randomised, double-blind, placebo-controlled, first-in-human phase 1 trial (SITU-SAFE) was conducted at the Phase 1 Unit, Uppsala University Hospital, Sweden, in 240 induced skin wounds in 36 healthy volunteers. The study included a treatment phase followed by an assessment phase running up to 6 weeks after last dose and an ongoing 5year long-term follow up. The results presented in this paper were collected between September 20th, 2019 and October 20th, 2021 include results up to 13 months, i.e. 12 months follow up after last dose in the MAD part. The primary objective was to determine the safety and tolerability profile, whereas other objectives included assessments of clinical and biologic effects on wound healing, as well as presence and biodistribution of ILP100-Topical. ILP100-Topical consists of L. reuteri R2LC genetically modified with the pSIP_CXCL12-α plasmid to express CXCL12-α, hereunder referred to as CXCL12, following induction by the activation peptide SppIP. 11,13,14 The ready-to-use drug product consists of L. reuteri R2LC carrying the pSIP_CXCL12 plasmid reconstituted with SppIP-containing buffer. As a control within each participant, placebo (SppIP-containing buffer), or saline (NaCl 0.9%) was used. The study was designed to comprise a single ascending dose (SAD) part of three cohorts, and a multiple ascending dose (MAD) part of another three cohorts, where safety confirmation of the SAD part preceded MAD initiation ( Supplementary Fig. S1).

Research in context
Evidence before this study We searched PubMed for original articles, meta-analyses, and systematic reviews published until April 25th, 2023, describing the role of CXCL12-α in regeneration search terms included but were not limited to: SDF-1, CXCL12, regeneration, wound. At the start of the study in 2019, therapeutic functions to promote wound healing had been successfully addressed in preclinical models using genetically modified cells or bacteria that delivered CXCL12 locally. There are to our knowledge no reports of CXCL12 being tested in human wounds prior to this study.

Added value of this study
To our knowledge, this study is the first to provide support for safety and effects on wound healing by the novel first-inclass drug candidate with therapeutic CXCL12-α (ILP100-Topical) in a blinded, randomised, and placebo-controlled clinical trial setting. In addition, we demonstrate that this newly designed biotechnological platform enables delivery of proteins with short half-life, e.g. chemokines such as CXCL12α, in a clinical use, and offers a novel approach for immunotherapies with local effects.

Implications of all the available evidence
No safety or tolerability issues were identified following treatment with ILP100-Topical to induced wounds. Clinical effect of accelerated wound healing was observed for the highest dose and when pooling data from all three multidose cohorts. The favourable safety profile and observed effect together support continued clinical development of ILP100-Topical for the treatment of difficult skin wounds in patients.

Articles
The studies were undertaken in accordance with Good Clinical Practice and the Declaration of Helsinki, and with approval of the Swedish Ethical Review Authority (Approval no. 2019-02802) and the Medical Product Agency in Uppsala, Sweden. Informed consent was obtained from the study individuals. The trial is registered in EudraCT (2019-000680-24).

Participants
Healthy male and female individuals aged 25-45 years who were willing to comply with the study procedures (experimental incision of 4 or 8 wounds in the SAD and MAD, respectively, equally distributed at the upper inner arms) and who have given written informed consent were considered eligible to participate in the study. Prior to consent, all individuals were given extensive information about the procedures and the potential risks with the study, such as punch biopsy procedure and risk of scarring. All individuals included had to understand and be willing to comply with study procedures. Individuals with a history of any bleeding disorder, including prolonged or habitual bleeding, individuals on blood-thinning medication or individuals with e.g. a tattoo or apparent skin abnormality on the upper inner arms were not included in the study, neither were pregnant or lactating women.

Randomisation and masking
The Investigational medicinal products (IMPs) were prepared by unblinded pharmacists, masked in order to maintain the blind, and administered topically in volumes of 50 μl per wound to blindfolded individuals. A computer-generated randomisation list (SAS Proc Plan, SAS Version 9.4, Institute, Inc., Cary, NC, USA) was kept by the randomiser in a sealed envelope until database lock.

Procedures
Enrolled individuals were admitted to the clinic on Day 1 for pre-dose safety assessments and full thickness wound punching (biopsy punch, 6 mm in diameter) on the ventral aspect of the upper arms (SAD: 2 wounds/ arm; MAD: 4 wounds/arm) following treatment of local anaesthesia (injected Xylocaine 10 mg/mL) and cleaning of the area (70% ethanol) (Supplementary Fig. S1). The wounds were photographed in a standardised setting before treatment on Day 1, and at all subsequent visits. For assessment of wound healing, the non-epithelialised wound area was measured by the IEs using ImageJ Software (U. S. National Institutes of Health, USA). To address exploratory objectives, wounds of the MAD part were scanned using a 3D spectroscopic scanner to evaluate scar area, scar volume and scar redness (Cherry Imaging, Yokneam, Israel, Supplementary methods) and blood perfusion of the wound bed and adjacent skin was measured (Laser Speckle Contract Analysis, LASCA; Perimed AB, Järfälla, Sweden, Supplementary Fig. S2 and Supplementary methods). 15,16 Wound biopsies were taken 48 h post-dosing in the SAD part for assessment of local mechanisms of action (Supplementary methods).
The SAD part of the study comprised of 3 sequential cohorts, each including 4 individuals with 2 experimentally induced wounds on each arm, in total 12 individuals and 48 wounds. For each individual, a single dose of ILP100-Topical (5 × 10 4 , 5 × 10 7 or 1 × 10 9 CFU/ cm 2 wound area in cohort 1, cohort 2 and cohort 3, respectively) and placebo were randomised to 2 wounds on the left arm and 2 wounds on the right arm, in a 1:1 ratio.
The MAD part comprised of 3 sequential cohorts, each including 8 individuals with 4 experimentally induced wounds on each arm, in total 24 individuals and 192 wounds. The IMP was randomised in a 4:2:2 ratio, with ILP100-Topical (cohort 1: 5 × 10 5 CFU/cm 2 , cohort 2: 5 × 10 7 CFU/cm 2 and cohort 3: 1 × 10 9 CFU/ cm 2 ) to 4 wounds on left or right arm, and placebo or saline to 2 wounds each on the arm on which wounds did not receive ILP100-Topical. Saline was used as a control to assess the potential effect on wound healing by the SppIP-containing buffer in placebo. Each wound was administered with repeated doses of IMP on Day 1, 2 and 3, followed by 3 times a week over the course of 3 weeks (10 doses in total).

Outcomes
Clinical safety assessments were performed at visits and included adverse events (AEs), clinical laboratory parameters, vital signs, ECG, physical examination, local tolerability reactions, formation of anti-CXCL12 antibodies (ADA, supplementary methods), systemic exposure of CXCL12 in plasma (Supplementary methods), as well as presence of L. reuteri R2LC containing the pSIP_CXCL12 on the skin surrounding the wound, blood, and faeces (Supplementary methods).
Tolerability, clinical and biologic effects were assessed at each visit (SAD: Day 1, 2, 3, 7, 14, and at 6 weeks, 3 months and 12 months from start of treatment at Day 1; MAD: Day 1,2,3,5,8,10,12,15,17,19,21,32, and at 6 weeks, 3 months and 12 months after last dose at Day 19). All assessments were blinded and occurred by on-site visual inspections of the wounds by the Principal Investigator (or co-investigator), as well as off-site by traceable evaluation and detailed wound area measurements from 2D photographs of all wounds by 3 Independent Evaluators (IEs) with expertise in wound healing. Tolerability was graded 0-3 according to predefined criteria based on wound appearance (wound and wound edge inflammation, surrounding skin inflammation, haemorrhage, presence of exudate, slough or necrotic tissue, granulation tissue, or hypergranulation). For the clinical effect on wound healing, a wound was defined as healed when the wound area was completely re-epithelialised and there were no dressing Articles www.thelancet.com Vol 60 June, 2023 requirements, and if the assessments by one or more IEs deviated more than two steps on the 4-grade scale, the three IEs assembled to adjudicate the definitive grade. In addition, 3D scans were used to assess changes in scar size, while analyses of the mechanism of action included blood flow measurements (MAD part only) and molecular changes by histology and local CXCL12 levels by ELISA in the wound biopsies (SAD part only).

Statistical analysis
The sample size was considered sufficient to provide adequate information for the primary and related safety and tolerability objectives. For detailed description about statistical analysis, please see supplementary methods. In the post-hoc analyses of the biologic and clinical effects on wound healing, Fisher's exact test and the Mann-Whitney test were used for comparing the different treatment groups for proportion healed wounds and average time to first registered healing. All descriptive summaries and pre-defined statistical analyses were performed using SAS Version 9.4 (SAS Institute Inc., Cary, NC, USA). Post-hoc analyses were performed using StatXact Version 11.1.0 (Cytel Inc., Waltham, MA, USA), SAS Version 9.4, and GraphPad Prism 9.1.1.225 (GraphPad Software, San Diego, CA, USA).

Role of the funding source
Ilya Pharma AB is the Sponsor of the study fulfilling all sponsor responsibilities. The trial was in part supported by a grant from the European Commission, H2020 SME Instrument Phase II (#804438) and by Knut and Alice Wallenberg foundation.
Study Sponsor was responsible for the study design, analysis of data from 3D scanning and LASCA measurements, and decision to publish the data. Study report and data interpretation (except for 3D scanning and LASCA measurements) was performed by CRO and reviewed by the Sponsor.
EÖ, EV, AF, HLT, PD, SJ, NT, MÅ, ZM, LR, MJ, PF, PH, SR, and MP all had access to the dataset and accept responsibility for the decision to submit for publication.

Results
Thirty-six healthy study individuals between 25 and 45 years old were enrolled at a Phase 1 Unit at Uppsala University Hospital, Uppsala, Sweden between 20th of September 2019 and 1st of October 2020 (Fig. 1). Baseline characteristics and demographics of the individuals are summarised in Supplementary Tables S1 and S2 for the SAD and MAD parts of the study, respectively.
The primary objective of the study was to determine the safety and tolerability profile. For all individuals, single-or multi-dosing of ILP100-Topical (1 and 10 administrations over 3 weeks, in the SAD and MAD, respectively) were considered safe and well-tolerated. No clinically significant changes from baseline of any parameters were detected during any visits. There were no serious adverse events or AEs leading to discontinuation from the study (Supplementary Tables S3 and S4). Overall, the AE profile of wounds treated with ILP100-Topical was comparable to that of wounds treated with placebo or saline (Table 1). For all cohorts, L. reuteri R2LC containing pSIP_CXCL12 was only identified on the skin surrounding wounds 1-2 days after treatment, no colonisation occurred, and L. reuteri R2LC containing pSIP_CXCL12 was not detected in blood or faeces at any time point. In addition, circulating levels of CXCL12 were not increased after single-or multi-dosing, and ADAs against CXCL12 could not be detected at any time point.
In both SAD and MAD, transient inflammation of the wound and surrounding skin was observed to a higher degree for the highest ILP100-Topical levels (Supplementary Tables S5 and S6), while the prevalence of wound infections was similar between saline, placebo, and ILP100-Topical treated wounds (Table 1). Treatment with ILP100-Topical was associated with increased exudation in the two lowest doses and in the highest dose to the amount of slough/necrotic tissue, as assessed by the IEs, but not according to the Investigators (Supplementary Tables S8 and S9 and data not included). There were no evident associations between the amount of, granulation tissue, haemorrhage or hypergranulation between the different treatments in either SAD or MAD (Supplementary Tables S8, S10, and S11). Irrespective of treatment in cohort 1 in the MAD part, the Investigators reported eczema and inflammation of the skin in contact with the dressing (Table 1), which resulted in discontinuation of treatment of in total 28 wounds (Supplementary Table S4). The dressing type was therefore changed for cohort 2 and 3. ILP100 treatment did not increase wound rupture, as this was only reported for scars of two placebo-treated wounds.
Secondary objectives included assessments of clinical and biologic effects on wound healing. No differences in wound healing were detected between the saline-or placebo-treated wounds by either Investigators or IEs, and saline-and placebo-treatment were therefore pooled. The Investigators' assessments did not show any difference in wound healing between treatment groups. In the MAD part, the IEs' assessments revealed treatment-related differences in wound healing at Days 32 in cohort 1, and at Days 19 and 21 (Day 32, p = 0.058) in cohort 3, where higher proportions of wounds treated with ILP100-Topical were assessed as healed by all three IEs compared to control-treated wounds ( Fig. 2A). A pooled analysis of all cohorts showed that ILP100-Topical significantly improved wound healing compared to controls (p = 0.020), as 76% (73/96) of the ILP100-Topical treated wounds were healed at or prior to  Day 32, as assessed by all IEs, compared to 59% (57/96) of control wounds ( Fig. 2A). Further, when all doses/ cohorts were pooled, the time to first registration of healed by all three IEs was on average shortened by 6 days by ILP100-Topical (p = 0.039) compared to controls. For the highest ILP100-Topical dose group, the time to first registration of healed was 10 days faster compared to controls (p = 0.0046, Fig. 2B). Similar results for time to wound healing and the proportion of healed wounds were obtained with paired statistical methods (data not included).
Irrespective of treatment, blood perfusion of the wound bed peaked at Day 8 (Supplementary Table S12), whereas wound edge perfusion decreased over time as the wounds gradually healed (Table 2) with the exception for cohort 1 where dressing-induced eczema and skin inflammation were reported (Table 1). Treatment with ILP100-Topical was found to increase wound edge blood perfusion dose dependently at Day 2 when compared to the control-treated wounds of cohorts 2 and 3, but not at Day 8 or 15 ( Table 2).
Immunohistochemistry of wound biopsies from the SAD wounds revealed increased numbers by 59% of CXCL12 + cells in the wound edge dermis by the highest dose of ILP100-Topical (1018 ± 134 vs 1623 ± 315 for control and ILP100-treated wounds, respectively) ( Table 3, Supplementary Fig. S3). No differences were detected for CXCL12 levels within tissue (Supplementary  Table S13).
Scar formation was assessed as normal for all healed wounds at all visits. The 3D spectroscopic scanning revealed no difference in scar areas between cohorts or treatments (Supplementary Table S14), while the sensitivity of scar volume scans did not allow for comparisons between treatment groups (Supplementary Tables S15 and S16). Scar redness normalised to skin colour was also assessed, but no differences between treatments could be detected (Supplementary Tables S17 and S18).

Discussion
In this first-in-human trial, topical application of the first-in-class drug candidate ILP100-Topical was suggested to be safe and well-tolerated. In addition, multiple doses of ILP100-Topical supported clinical efficacy on wound healing, as demonstrated by a larger proportion of healed wounds from Day 19 and shortened time to first registered healing. Thus, genetically modified L. reuteri R2LC engineered to deliver and stabilise CXCL12 was suggested to be safe and effective in accelerating healing of induced wounds.
Therapeutic means to support healing has recently been focusing on altering the wound microenvironment  Infections and infestations 1 (3.1%) Percentages are based on the number of wounds in the study period included in the full analysis set, n, number of wounds; m, number of events. Pre-treatments are not included.  by topical application of growth factors, plasma-derived products or cell therapies. [17][18][19] These strategies are often hampered by restricted access of administered cells to wound tissue, and by limited bioavailability of therapeutic proteins due to high levels of proteases present in wounds. Another disruptive and recently recognised approach is to use genetically modified bacteria to deliver endogenous proteins to wounds. 20 So far, two attempts have been reported to accelerate healing in mouse models: Lactococcus lactis expressing FGF2, CSF1, and IL-4 (AUP-1602-C) currently tested in a firsthuman trial (NCT04281992), and the herein investigated ILP100-Topical, L. reuteri R2LC expressing CXCL12. 11,21 In addition to the onsite bacterial production, the lactic acid produced by L. reuteri R2LC was demonstrated to reduce CXCL12 degradation within the wound, and thereby further boosting the CXCL12induced tissue restorative functions of macrophages. 11 Accelerated healing of wounds by ILP100-Topical was also confirmed in minipigs. 12 For new modalities, trial design capturing drugspecific characteristics are vital for continued clinical development. The present trial was designed to allow independent evaluations of wounds, reduce the number of individuals and overcome interindividual variability by having wounds treated with ILP100-Topical, placebo and saline in the same participant. The individuals were closely monitored using an extensive set of safety and tolerability assessments, and all wounds were imaged for subsequent off-site, unbiased, high-resolution, and traceable analyses, in addition to the on-site assessments. Notably, no clinically significant deviations from baseline were detected when safety and tolerability were assessed, and no serious AEs were recorded. Treatment of acute wounds with ILP100-Topical was therefore demonstrated to be both safe and well-tolerated at all timepoints and doses tested.
Complete wound healing is the regulatory endpoint considered to be the most clinically meaningful by FDA. In this study, wound healing was assessed by blinded and fully traceable, off-site analyses of highresolution wound images. All three IEs found that a higher proportion of wounds treated with the highest dose of ILP100-Topcial were healed from Day 19. Further, the time to first registration of complete healing was shortened by 10 days following repeated ILP100-Topical treatment with the highest dose. These results are indeed clinically very relevant given that 1-2 days of accelerated healing in acute wounds or healing of 10-15% more non-healing ulcers in patients with diabetes compared to standard of care is regarded as clinically meaningful and suffice the requirements for marketing authorisation by regulatory authorities. 22 To increase the probability of capturing AEs and effects on wound healing in this first-in-human trial, we combined the clinical assessment of the wounds with objective, explorative techniques measuring local blood perfusion and scar formation. These different techniques together generated more than 100 000 data points analysed in a blinded manner. While the small size of the scars precluded comparisons between treatments, a transient and dose-dependent hyperaemia around the ILP100-Topical-treated wounds were detected at early time points. Together with the observed accelerated healing and limited number of transient inflammation-related AEs, this likely reflects biologic effects of the treatment, rather than inflammatory response to bacteria. Thus, the obtained results support continued clinical development of ILP100-Topical for the treatment of difficult-to-heal skin wounds in patients. In fact, two phase 2 trials investigating ILP100-Topical as treatment in different wound types is now approved by European and US health authorities.
three cohorts of the MAD part, as well as for pooled cohorts. Mann-Whitney test *p ≤ 0.05, **p ≤ 0.01. Average time, error bars indicate standard error of the mean (SEM). Wounds with missing timepoint of wound healing or not judged as healed by the end of the study has been imputed as healed after 61 days being the next timepoint of assessment. Control group includes saline-or placebo-treated wounds. The MAD cohorts treated with 5 × 10 5 CFU/cm 2 wound area (cohort 1), 5 × 10 7 CFU/cm 2 (cohort 2) and 1 × 10 9 CFU/cm 2 (cohort 3), respectively, were assessed for wound edge perfusion at visits

Articles
Limitations of this study include the single-centre design, the rather small number of study individuals and that different investigators were involved in clinical assessments. Changes in the investigator team during MAD cohort 2 might have influenced the wounding procedures and thereby explain inconsistent results compared to other cohorts. In addition, the individuals included in this study were all healthy, non-obese, and under the age of 45, and is thereby not predisposed for these factors associated with impaired or complex wound healing. Hence, while ILP100-Topical in this study show results supporting an accelerated wound healing in otherwise healthy patients (eg in traumarelated wounds), the effect might not be directly translatable to a patient population exhibiting risk factors for delayed wound healing. As a natural next step in the clinical development efficacy is already being investigated in different well-defined patient populations with pathologies linked to impaired wound healing. Strengths include its design allowing large numbers of wounds, minimal bias as wounds treated with active and control treatment in the same individuals reduce the risk for factors influencing wound healing in different treatment groups, as well as high comparability between treatments for tolerability and effects on healing. This is especially important in a First-in-Human study with few study individuals and at the same time allows for a smaller samples size with fewer individuals exposed to an experimental investigational product in early clinical development. The well-being of the study individuals was thoroughly considered, and each study individual was informed about the study procedures and risks for scar formation before giving consent. Only individuals able to fully understand the study information and comply with the protocol procedures were considered for the study. Further, each study individual fulfilled the inclusion criteria and did not present any of the exclusion criteria including conditions associated with abnormal scar formation and other physical risks, but not mental illness risks. The latter was not considered necessary to evaluate specifically, given that the individuals were informed and accepted the study risk and were assessed for their overall eligibility for participation. Another strength is the wound assessments from high-resolution images, which allows blinded, detailed analyses of both tolerability and wound healing. In conclusion, the favourable safety profile together with the clinical and biologic effects on wound healing support continued clinical development of ILP100-Topical for the treatment of complicated, nonhealing wounds in patients. In addition, the study demonstrates that genetically modified bacteria is a new modality enabling the use of short-lived proteins, such as CXCL12, as therapeutics.

List of Tables
Supp.

Study procedures
The SAD part of the study included 6 visits for treatment and initial follow-up (Day 1 to Day 14). In the MAD part, there were 12 visits for treatment and initial follow-up (Day 1 to Day 32). All individuals are part of a 5-year longterm follow-up. This report includes results from visits up to 6 weeks, and at 3, 6 and 12 months after the last dose was administered, referred to as 2, 4, 7 and 13 months indicating time post wound induction. After the initial screening visit, individuals fulfilling the study criteria were enrolled.
In SAD, separate adhesive transparent film dressings were used to isolate the wounds from each other. In cohort 1 in the MAD the same dressing was applied throughout the treatment period (Day 1 to Day 19). With repeated administrations of IMP occlusive film dressing caused eczema on the skin surrounding the wounds in the majority of study participants. In cohorts 2 and 3 of the MAD part, the dressing was therefore changed and covered with adhesive, transparent film during 48 hours after the first and second IMP application. From Day 3 and onwards, the wounds were treated with IMP and then covered with adhesive, transparent film for 1 hour only. Thereafter the film was removed, the wounds were allowed to air dry and were then be covered with non-occlusive dressing in accordance with standard wound care procedures. Each wound was dressed until healed.

3D spectroscopic scanner evaluation
In order to understand the precision and the limitations of the 3D spectroscopic scanner (Cherry Imaging, Yokneam, Israel), an evaluation of the scanner was performed using scars scanned in the SAD part. The scar volume and area evaluation were performed 2 months post wound induction, where four scars from the same subject were scanned consecutively for four times in order to assess inter-scan variability. To assess intra-scan variability, each scar was measured five times using the Cherry Imaging software. The scars used for this evaluation were very small with areas and volumes ranging from 23-26 mm 2 and 0.7-1.6 mm 3 , respectively, and the results are presented in Supplementary  table 15. Evaluation of the measurements of the redness of scars was performed in a similar manner where wounds or skin areas with a redness score of 0.1 to 0.9 were measured repeatedly. For the inter-scan variability 10 areas were used. In total, 20 areas per wound were included in the intra-scan variability assessment, and the defined area within the scan was repeatedly measured five times using the Cherry Imaging software. The results are presented in Supplementary table 17.

Study outcomes
CXCL12 levels in human plasma were analysed using ELISA according to the manufacturer's instructions (Human CXCL12/SDF-1a Quantikine ELISA kit and Quantikine Immunoassay Control Group 3, R&D Systems, Minneapolis, MN, USA). To determine the presence of ADAs, human plasma samples were analysed using a GLP-validated electrochemiluminescent immunoassay (ECLIA). Presence of L. reuteri R2LC containing the pSIP_CXCL12 plasmid was analysed in faeces, blood samples and swabs of the area surrounding the wounds by bacterial culturing. PCR and sequencing were used in the occurrence of bacterial culture colonies.
In the SAD part, one placebo-treated wound and one ILP100-Topcial-treated wound were biopsied again at 48 hrs post wounding with an 8 mm in diameter biopsy punch. The biopsy was split in two halves, one was used for histology and one for analysis of tissue CXCL12 by ELISA (Quantikine ELISA Human CXCL12 / SDF-1α Immunoassay, R&D Systems, Minneapolis, MN, USA). The tissue saved for histology was paraffin embedded and stained for CXCL12 (NSJ Bioreagents, RQ4559).
The 3D scans of scar area, volume and pigmentation were analysed (Trace software, Cherry Imaging), and validated (Supplementary Methods and Supplementary Tables 1-2). Blood flow was recorded in an area of 5x10 cm around 2 wounds at a time during at least 2 minutes. Reference perfusion was measured in an area remote from the wound (Supplementary Figure 2), whereas wound edge perfusion was measured in the surrounding skin within 5 mm from the wound border. Blood perfusion of the wound edge is reported as delta perfusion units (dPFU; wound edge perfusion subtracted by reference perfusion). In the SAD part of the study, local mechanism of action was assessed by histology of wound biopsies stained for CXCL12 (NSJ Bioreagents, RQ4559), as well as ELISA to measure total local CXCL12 levels in the wound and immediate surrounding tissue.

Statistical analysis
The statistical analyses for safety and clinical efficacy endpoints included all randomised individuals who received at least one dose of the IMP (Full analyses set; FAS).
No formal sample size calculations were performed for this first-in-human study with the primary objective to study safety and tolerability. The duration of the Treatment Period was selected as sufficiently long in order to assess the safety, PK/PD and preliminary efficacy of ILP100-Topical treatment. This is anticipated long enough to provide initial information about a clinical efficacy during treatment and sufficiently long to capture delayed AEs, a delayed onset of action and requirement of maintenance treatment versus a single dose.
Predefined statistical analyses included a mixed linear regression model for analyses of pairwise (left and right arm as well as wound position on the arm) treatment comparison of time to first registered wound healing and McNemar's paired test for proportion of healed wounds at each timepoint. However, at study design and regulatory approval, as well as at database lock, the wounds in the different treatments groups were considered to be most appropriately analysed as independent based on the influence from biologic parameters related wound healing associated with different use of the dominant and non-dominant hands. These parameters include blood circulation, muscle mass, activity, metabolism and structures of underlying muscles and other tissues, as well as mechanical impact on underlying tissues and abrasion of the skin. In the post-hoc analyses, the biologic and clinical effects were analysed using Fisher´s exact test and the Mann-Whitney test for comparing the different treatment groups for proportion healed wounds and average time to first registered healing. Since the IEs did not perform assessments after Day 32, and the Investigators assessed all wounds as healed on Day 61 (2 months) after wound induction, 61 days was imputed as the timepoint for healing if no earlier timepoint was registered for healing or for wounds with missing data. A safety review committee reviewed all safety and tolerability data throughout the treatment phase. Given the primary objective to assess the safety and tolerability, and the hypothesis-testing nature of the biologic and clinical assessments of wound healing, no adjustments for multiplicity was made.

Supplementary tables
Supplementary     One study individual discontinued treatment due to wound infection after 4 doses of ILP100-Topical, 4 doses of placebo, and 6 doses of saline. Another study individual discontinued treatment on placebo and saline treated wounds due to wound site eczema after receiving 4 doses. b.
One study individual did not perform day 12 to 19 due to upper respiratory tract infection, received 6 doses of ILP100-Topical, placebo, and saline. c.
One study individual did not perform day 17 due to diarrhea. This study individual received 9 doses of ILP100-Topical, placebo and saline.